OCCURRENCE OF SURFACE RUNOFF FROM GEORGIA PIEDMONT PLANTED AND MATURE FORESTS: THREE CASE STUDIES by JOAN V. SCHNEIER (Under the Direction of C. Rhett Jackson) ABSTRACT This study evaluated overland flow exiting two clearcut, bedded, and planted areas plus an uncut reference area. Runoff cups were established at 30 m intervals along the boundary of streamside management zones (SMZs). Absence or presence of runoff was tallied for events exceeding 13 mm and concentrated flow tracks (CFTs) were grab sampled during large storms. Mean runoff cup response was 9.7% from the plantations and 2.2% for the reference. Cup responsiveness and bare ground decreased from the first to the second year after planting. Response frequency was best correlated to rainfall factors “R” from the Revised Universal Soil Loss Equation, total storm depth, and 24- hour and 6-hour maximum intensities. Runoff locations were well distributed and some CFTs fully penetrated the SMZ. Mean plantation concentrations of dissolved nitrates, dissolved phosphates and suspended solids were 2.1, 0.21, 54 mg/L the first year, and 0.1, 0.12, 36 mg/L the second year, respectively. INDEX WORDS: Overland flow, surface runoff, concentrated flow tracks, forestry water quality, streamside management zones, buffer strips, sediment, nitrates, phosphates
142
Embed
OCCURRENCE OF SURFACE RUNOFF FROM GEORGIA …getd.galib.uga.edu/public/schneier_joan_v_200605...OCCURRENCE OF SURFACE RUNOFF FROM GEORGIA PIEDMONT PLANTED AND MATURE FORESTS: THREE
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
OCCURRENCE OF SURFACE RUNOFF FROM GEORGIA PIEDMONT PLANTED
AND MATURE FORESTS: THREE CASE STUDIES
by
JOAN V. SCHNEIER
(Under the Direction of C. Rhett Jackson)
ABSTRACT
This study evaluated overland flow exiting two clearcut, bedded, and planted
areas plus an uncut reference area. Runoff cups were established at 30 m intervals along
the boundary of streamside management zones (SMZs). Absence or presence of runoff
was tallied for events exceeding 13 mm and concentrated flow tracks (CFTs) were grab
sampled during large storms. Mean runoff cup response was 9.7% from the plantations
and 2.2% for the reference. Cup responsiveness and bare ground decreased from the first
to the second year after planting. Response frequency was best correlated to rainfall
factors “R” from the Revised Universal Soil Loss Equation, total storm depth, and 24-
hour and 6-hour maximum intensities. Runoff locations were well distributed and some
CFTs fully penetrated the SMZ. Mean plantation concentrations of dissolved nitrates,
dissolved phosphates and suspended solids were 2.1, 0.21, 54 mg/L the first year, and
0.1, 0.12, 36 mg/L the second year, respectively.
INDEX WORDS: Overland flow, surface runoff, concentrated flow tracks, forestry
water quality, streamside management zones, buffer strips, sediment, nitrates, phosphates
OCCURRENCE OF SURFACE RUNOFF FROM GEORGIA PIEDMONT PLANTED
AND MATURE FORESTS: THREE CASE STUDIES
by
JOAN V. SCHNEIER
B.S., Mississippi State University, 2001
A Thesis Submitted to the Graduate Faculty of The University of Georgia in Partial
S. CA 20 Te-SA LS 1.1 19 Bu 8-16 10 K 20-50 2W.Africa 4 Tr-SH 1077 S,LS Ag 150 60 <3.9 p H 3East US 23 Te-Hu 1372 SL 36 med 99 <3 p,e 4
VT 30-100 Te-Hu SL 0.2 pasture I >80 8 a S,R 5Japan mtn 63 Te-Hu 1459 volc.ash 2.5 13 30 K 22 1.3 p S 6
Vietnam mtn 0-173 Te-Hu 1800 SCL 57, 85 30,10 K 63 .01 p H 7Vietnam mtn Te-Hu 1800 SCL AF 57, 85 30,11 K 28 8.8 p H 7Vietnam mtn Te-Hu 1800 SCL 0 Rd 57, 85 30,12 K 7 47.2 p H 7Panama 5-10 Tr-RF 2600 25 49.9 f 8Panama 5-10 Tr-RF 2600 40 21.3 f 8Panama 20 Tr-RF 2600 SiC 10 160 5 6.9 p,e S 9Brazil low Tr-Rf 2870 SC,SCL 23 2.8 a S 10Peru foot. Tr-RF 3300 0.75 23, 52 60, 5 I 317 27 f S,R 11
1 Not a comprehensive list. 2 Some of the data were estimated from graphs or figures. 3 Climate- Te=temperate, SA= semi-arid, Hu= humid, RF= rain forest. Cover- Past.= pasture4 Soils- C= clay, L= loam, S= sand, Si = silt. 5 Disturbance- Bu= severe burn, Ag= upland rice, AF= abandoned field, Rd= road6 Intensity (x) = maximum intensity for x minutes. These could be mean,median, or absolute maximums.7 Ksat (K) and Infiltration (I) could be field or lab values. Infiltration could be final or an intermediate value.8 Overland flow amount-a=area, e=estimated, f=frequency, p=precip depth.Type-H= Hortonian, R=return flow, S=saturation overland flow9 Type- H= Hortonian, R= return flow, S= saturation overland flow10 Studies:
1 Heede 1987 5 Dunne & Black 1970a 9 Dietrich et al 19822 Wohlgemuth et al. 2001 6 Sidle et al. 2000 10 Lesack 19933 van de Giesen et al. 2000 7 Ziegler et al. 2004 11 Elsenbeer &Vertessy 20004 Hewlett and Hibbert 1967 8 Godsey et al. 2004
-----Overland Flow---------Maximum Rain----
(mm/hr) Infiltration7
Ksat or
30
Table 1.2. Presence of overland flow in southeastern US regeneration studies1
Bare OverlandLocation2 Slope Area Soil Cut3 Regeneration4 Flow Author Year
(%) (ha) (%)UCP TX 4-25 2-3 3 n/a n/a No Blackburn et al. 1986
16 CC C,Bu,HP No57 CC S,P,Bu,HP Yes
UCP TX plot CC CC,Bu Yes Field et al. 2003CC CC Yes
UCP MS 38 <1 37 CC C,B,HP? Yes Beasley 197953 CC S,P,Bu,HP? Yes69 CC S,P,Be,HP? Yes
UCP MS 0.7 CC n/a Yes McClurkin et al. 1987Th n/a No
P NC 4 0.1 18 CC S,Pi,D,HP Yes Pye and Vitousek 19853 0.1 32 CC S,Pi,D,Ch,HP Yes
P GA 50 CC 2C,MP Yes Hewlett 1978Hewlett et al. 1984
P AL 10 plot CC S,Pi,Be,MP Yes Grace 2004CC MP Yes
Mtn SC plot 7 CC Bu Yes Robichaud and Waldrop 199463 CC Bu Yes
Mtn OK 57 1-5 15 CC Cr,Bu,R,HP Yes Miller 1984Mtn WV 35 CC n/a Yes Patric 1980
Table 1.3. Citations and locations for southeastern pine regeneration stormwater studies for small streams and runoff plots
Study No. Location Data Source Author Year1 Upper Coastal Plain AR Ephemerals Beasley et al. 19862 Lower Coastal Plain AR Intermittent Beasley and Granillo 19883 Upper Coastal Plain TX Ephemerals Blackburn et al. 19864 Upper Coastal Plain TX Ephemerals Blackburn and Wood 19905 Piedmont SC Ephemerals Douglass and Van Lear 19836 Upper Coastal Plain TX Plots Field et al. 20037 Upper Coastal Plain TX Plots Field et al. 20058 Piedmont VA Ephemerals Fox et al. 19869 Piedmont AL Plots Grace 2004
10 Piedmont GA Perennials Hewlett et al. 198411 Upper Coastal Plain TN Plots McLurkin et al. 198512 Upper Coastal Plain MS Plots McLurkin et al. 198713 Eastern US1 Perennials McDowell and Omernik 197714 Eastern US1 Perennials Patric et al. 198415 Piedmont NC Plots Pye and Vitousek 198516 Mountains SC Plots Robichaud and Waldrop 199417 Upper Coastal Plain MS2 Ephemerals, Swales Schreiber et al. 198018 Upper Coastal Plain MS Unspecified headwater Ursic 199119 Piedmont SC Ephemerals Van Lear et al. 198520 Piedmont GA Swales Ward and Jackson 2004
1 Regional average, all stages of rotation2 36 years old
32
Figure 1.1.Rainfall and hillslope hydrology processes. From Jackson (in press) and Atkinson (1978)
33
CHAPTER 2
METHODS
Study Sites
Three study sites were established in Greene County, Georgia, about 40 km (25
miles) southeast of Athens, latitude 330 40’, longitude, 830 15’ (Figure 2.1). The Lewis
and Vanir tracts were portions of recently regenerated Plum Creek Timber Company pine
stands, and the Watson Springs tract was a mature pine-hardwood forest managed by the
UGA Warnell School of Forestry and Natural Resources. The upland areas of the tracts
were 82 ha (203 ac) for Lewis, 27 ha (67 ac) for Vanir, and 30 ha (74 ac) for the Watson
Springs reference site.
Elevations ranged from 139 to 197 meters (456 – 646 ft), and slopes from 0 - 23 %
(Table 2.1). Median slopes were 6.7 % on the two plantations and 8.2 % at the reference
site. Topsoils textures at Lewis were 93% sandy loam (mostly Cecil and Pacolet series),
88% sandy loam or loamy sand at Vanir (mostly Cecil, Pacolet, and Rion series), and
75% sandy loam or loamy sand at Watson Springs (mostly Louisberg, Pacolet, Vance,
and Cecil series). The soils were mostly well-drained Ultisols, although Lewis had wet
areas, concentrated on the south and west sides (Figures 2.2 -2.4). The reference site was
steeper than the plantations but also had sandier subsoils and lower erodibility as shown
by the USLE “K” values. Stream and gully conditions and also archaeological artifacts
indicate all sites were intensively row cropped in the nineteenth and early twentieth
centuries, ass is common in the Piedmont. before modern conservation practices (Trimble
1974). Aerial photos from 1942 showed that about 30% of Lewis, 30% of Vanir, and 16
34
% of the reference site were in crops, pastures, or open canopy areas, sometimes with
evidence of terraces. Major portions of the forested areas appeared understocked.
Agriculture created numerous relict gullies which were reasonably stable when covered
by forest and litter but subject to reactivation when disturbed (e.g. Hewlett 1978;
Rivenbark and Jackson 2004). Even when stable, they were sometimes observed
collecting and channeling water during large storms.
The two regeneration sites were selected from 12 candidates based on presence of
streams with SMZ buffers, typical industry treatments, topographic variation, access,
size, and lack of non-forestry perturbations. Only two were chosen due to the labor
intensive nature of the research. On each site, 45-55 year old pines of the previous
rotation were clearcut to commercial specifications in 2001 (Table 2.2). The sites were
aerial sprayed with 6.7 kg/ha (6 lbs/ac) hexazinone (Velpar® ULW) for hardwood
control in May of 2002 without burning. Areas of heavy debris were spot piled in August.
The entire tract was then ripped to a depth of 0.5 m (18 inches), and simultaneously low
bedded with a combination plow. The beds were planted in loblolly pines in January
2003 at 1579-1678 trees/ha (635-679 trees per acre) and had no subsequent treatments
(Joelle Hairell and Grady Britt, personal communication). An excellent job of contour
bedding was done. The SMZs below the plantations were only minimally disturbed
during logging, which left basal areas of about 24 m2 / ha (105 ft2 / ac), canopy shading of
over 87%, and bare ground of less than 3% (Table 2.3). This far exceeded minimal
Georgia BMP requirements of 11 m2 (50 ft2) of basal area or 50% canopy coverage for
SMZs (Ga. EPD et al 1999).
35
The reference site was established in about 70- year mature timber in February 2004
at the Watson Springs Research Forest, about 4 km north of the Lewis tract. This tract
was acquired by University of Georgia in 1933 (Dustin Thompson, personal
communication) following the burning of a resort in 1930 and the consequent
abandonment of the small support town (Roper 1996). The land regenerated naturally to
mixed pine and hardwood, with the pine stands thinned and salvaged infrequently and
burned more regularly. Most of this study site was prescribed burned shortly after study
setup, resulting in a patchy, thin litter layer, and reactivation of some gullies. The
reference area had 25 sq m/ha (110 ft2 / ac) timber basal area, 10% bare ground post-burn,
and canopy shading of 97%, in the SMZ.
Rainfall Measurements
At each site, rainfall was measured with an Onset RG-2 tipping bucket rain gauge
with a HOBO event recorder, which recorded every 0.03 cm (0.01 inch of rain). The
tipping bucket gage readings were within 5 % of standard gauge (True-Check brand)
readings for the same sites. Freezing rain and snow events practically never occurred.
However, several months of data were lost on the plantations due to malfunctions, launch
failures, and plugging of the bucket outlet by seeds. During periods of tipping gage
malfunction, the standard gage totals were distributed using data from the USGS gauge at
Penfield (#02218300) 2 km (1 mile) north of one site and/or data from the other sites.
Storm data for February 2004 prior to installation at the reference site were interpolated
by averaging the data from Lewis and USGS Penfield, since winter storms tend to have
uniform rainfall over a wide area and the reference was between these two sites. Only
rain depth values were adjusted, with no attempt made to reconstruct intensity or
36
duration, since these are highly localized. In spite of these problems, most of the tipping
bucket data were good for most of the study. No other nearby published rain data were
found during an internet search.
The spring and summer of 2003 were wetter than normal and the following fall,
winter and spring drier than normal. The summer of 2004 was wetter than normal with
the final study month of September far exceeding the norm (Tables 2.4, 2.5, Figure 2.5).
Runoff Data Collection
Runoff cups (Dunne et al. 1975) were placed in a single line at 30 meter taped
intervals along the top edge of the SMZ at the plantations and along the simulated edge
for the reference site. Heavy logging debris or roots occasionally necessitated slight
placement adjustments. A 9 cm (3.5 inch) diameter bulb planter was used to drill a
smooth-sided hole in the ground and a Solo 270 ml (nine ounce) plastic beverage cup
placed inside the hole with the lip level with the ground. Two 30 cm (12 inch) stakes
were driven on the uphill side of the cup and a 900 cm2 (144 in2) piece of a roofing
shingle tacked onto the stakes covered the cup from rain (Figure 2.6). A total of 271
runoff cups were installed, 123 cups at Lewis, 69 at Vanir, and 79 on the reference site
(Figures 2.2 -2.4). The path for checking line was established downhill of, or connecting,
the cups except in a few places where areas of heavy debris or gully hazards necessitated
walking uphill. In those cases, an effort was made never to step directly in front of a cup,
especially during wet weather, to avoid compacting the soil.
After each rainfall in excess of 13 mm (0.5 inches), as shown on the standard
gauge, cups were checked for absence or presence of water, emptied, and the setup reset
37
and repaired as needed. In practice, cups were often checked after 9 cm (0.35 in) events.
Even in the absence of rainfall, the cups required inspection every two weeks for minor
repairs. Cups were usually checked within 48 hours post-storms. The two planted sites
were each checked about 60 times over the 20 month study period, and the control site 27
times over eight months (Tables 2.4, 2.5).
Early in the study a bimodal distribution of volumes was observed, resulting in a
rough classification system. Volumes of water in the cups were estimated by quarters
plus two additional categories of “trace” and “submerged” were added (Table 2.6).
“Trace” was just a few drops of water insufficient to cover the bottom of the cup, while
submerged meant the cup was underwater in a puddle or ephemeral flow and the cup
could not be emptied and/or replaced in the hole.
Several problems were caused by either too little or too much water. “Trace”
tallies (<5 ml) were thrown out due to possible confounding effects of raindrop splash,
condensation, blowing rain, and drip from the underside of shingles. However, at least
some of these small cup volumes could have actually been due to runoff. Excess water
sometimes caused full or partial displacement of cups from holes (“risers”), particularly
in wet areas of Lewis. Other causes of elevated cups were: filling of the hole under the
cup with mud, dirt, or roots, animal activity, or cups sticking to the glue on the shingles
in hot weather. The most consistent “risers” on the south end of the Lewis tract were
eventually weighted with 2300 g (5oz) of fishing sinkers. While the glue problem was not
solved, replacement shingles were installed gravel side up, leaving the largest glue patch
on top of the shingle. Since the vast majority of risers seemed to be caused by high
ground water in variable source areas, this tally was included in the frequency data.
38
Submerged cups were counted until the area dried around them. The tally was thrown out
once at that point, since the contents could not be positively assigned to an individual
storm, they were then emptied and reverted to normal status.
Frequently occurring repairs were: tacks pulling through the shingles, loss of
shingle integrity over time, stake rotting or splitting, holes filling, and ant poisoning. In
most cases these issues were corrected before they became major problems. The tally was
thrown out for individual cups with serious problems, until fixed. The shingles evidently
provided attractive cover for ants (especially fire ants, Solenopsis spp.). When ant nest
building became a problem uphill of the cup, the cup setup was moved up or downhill by
a meter, perpendicular to the cup line. About 15% of the cups were moved during the
study, and ants were poisoned on nearly every trip in the summer. The immediate areas in
front of the cups were clipped during the first season to minimize canopy drip from
vegetation and brush, but trees (> 7.5 cm or 3 inch stump diameter) on the SMZ
boundary were left in place.
Grab Sampling and Collectors
A limited amount of water chemistry data (sediment, nitrates, and phosphates)
were obtained from grab sampling during six major events, which included five named
tropical storms (Table 2.6). Two storms were sampled in the summer of 2003, one in the
winter of 2004, and three in the summer of 2004. Sampling locations were generated
from a random number list of cups at each planted site. Runoff water from concentrated
flow tracks was located near each sample number, collected in an acid-washed Nalgene
bottle, and the location flagged for later mapping (Figures 2.7, 2.8). If no flow was found
39
in the area, puddles were grab sampled that had been recently flowing into the SMZ as
shown by matted vegetation. The samples were iced down at the truck, brought to the lab,
and filtered for sediment, using coarse and medium prefilters if necessary, ending with a
fine filter. The fine filters were Whatman 934-AH borosilicate fiberglass with 1.5 um
diameter pores, as specified in Standard Methods (Eaton et al. 1995). The filters were
weighed for Total Suspended Solids (TSS) and the filtrate analyzed for nitrates and
phosphates using a Hach DR890 Colorimeter. The concentrated flow tracks noted during
the storms were often not discernible afterwards due to small size, non-disturbance of
litter, and only temporary disturbance of vegetation. Runoff effects, frequency, duration,
and visibility of flow tracks decreased noticeably during the latter period of the study.
Seven metal collectors, as described by Franklin et al. (2001) and Sheridan et al.
(1996) were installed at Vanir during the summer of 2003 a little farther uphill than the
runoff cups to avoid effects of canopy drip. TSS data from nine storms was obtained from
them the following summer.
Cover Plots
All vegetation data were taken at the stand scale, and not linked to specific runoff
cups. Cover plots were taken at the end of the first and second growing season (Sept 2003
and August-September 2004) but before leaf drop (Figures 2.7 – 2.9). The two
plantations were surveyed using a line plot method by hand compass and pacing, similar
to standard timber cruising procedure. Categories were bare ground, gravel/rock, litter,
plant, and woody debris. Litter was defined as dead leaves, dead plants, or twigs under ½
inch (5 cm) in diameter, or small or nearly rotted bark flakes. Readings were made just
off the side of a two-meter pole laid on the ground. Nine readings were taken at each plot
40
at one meter intervals in a cross pattern (four in the line of travel, four at right angles,
and one at plot center). Other parameters measured were depth of lightly compressed
litter at plot center and tallest height of dominant vegetation (almost always herbaceous)
within a two meter radius of plot center.
Plots were also taken in the SMZ of the plantations. In addition to the plot system
described above, other data included timber basal area, midstory, and densiometer canopy
readings (Table 2.3). Basal area of pine and hardwood was taken with a 10 factor prism,
for trees greater than or equal to 13 cm (five inches) diameter at breast height (DBH) and
converted to square meters per hectare. Midstory was counted inside of a 1/1000 hectare
plot for stems from 0.1-13 cm at DBH. In both cases, all live woody (but not vines or
herbaceous) stems were counted. Two spherical densiometer readings of canopy cover
were taken at each plot. The directions of the readings were generated from a random
number list. Because of the linear and directionally erratic nature of the SMZ, the plots
followed the corridor with a fixed distance between plots. Directions were set at 10
degrees off the stream or SMZ edge and reset whenever a boundary was reached. The
entire SMZ was traversed in the area covered by the cups. A similar system was also used
on the reference area, except a cardinal direction line plot system was used for upland
plots, with zigzag plots in the SMZ (Table 2.3).
Mapping and GPS
Site features were mapped with a Trimble GeoExplorer 3 GPS unit with a
theoretical maximum Horizontal Dilution of Precision (HDOP) below three meters
(Table 2.7). Most features were mapped as points with the more important or problematic
41
points repeated 3-10 times for better accuracy. The cups were mapped during the dormant
season due to heavy canopy in the SMZ. Due to canopy interference, streams at Vanir
and the control site and cups at the control site were remapped with a Trimble ProXR
with an external antenna with a theoretical mean HDOP below 1.1 m.
All readings were differentially corrected. Features such as public roads near the
study site were obtained from the Georgia GIS Clearinghouse (http://gis1.state.ga.us),
county maps and 1999 Digital Ortho Quarter Quads (DOQQs). Soil maps of the planted
sites were obtained from the landowner and digitized in. NRCS (in Greensboro and
Athens) supplied ArcView shapefiles for the Watson Springs area, since the soil survey is
unpublished. They also supplied “K” values and other soils information. Decks and larger
skid trails were traced from large- scale (1:7920) post –logging aerial photos taken by the
landowner, scanned and digitized, rubber sheeting from the roads shown on the photos
and previous GPS work.
Gullies and larger active or inactive concentrated flow tracks that were near the
SMZ were also GPS mapped. Partway through the project, a switch was made from
mapping gullies with line features to mapping with points (usually 10 per position),
which gave more accurate positions. Relevant streams were point mapped in a similar
fashion to gullies. Downloading and clean up were done as soon as possible post field
work.
GPS features and other digitized data were converted into ArcView 3.2 shape files
used for generating maps. Frequently used extensions were Spatial Analyst and XTools.
Elevations and slopes used in site description were derived from Digital Elevation
42
Models (DEMs) downloaded from the Georgia GIS Clearinghouse, and converted to
raster using 30 meter cells.
Concentrated Flow Tracks (CFTs)
Measurements were taken on concentrated flow tracks (CFTs) which were still
active and which crossed into the SMZ, in July through August 2004 (second growing
season of the study). The SMZ boundary used in the plantations was the runoff cup line
or standline established by management. Since all the timber was standing at the
reference site, the minimum distances and slopes suggested by the forestry BMP manual
(Ga. EP Division et al. 1999) were used for the SMZ boundary (Table 2.8). These did not
always coincide with the runoff cups, which were visually estimated factoring in slope,
distance, and vegetation, according to common industry practice for establishing SMZs.
The classification system for CFTs was the same one used by Rivenbark and
Jackson. (2004) (Table 2.9), based on texture delivered to the stream. “Active” was
defined as having a visible path cut through the litter into the SMZ. Measurements, which
were more quantified in this study, included length and width of eroding areas and
ground cover in the channel, sidewall, contributing area, and nearby area. The same five
categories of ground cover used in the cover surveys were classified at 50 intersects
within a 0.5 meter (1.6 ft) by 0.25 meter (0.8 ft) PVC frame stung with wires. Slopes
were measured by clinometer. CFT width was measured at the flat part of the channel
bottom. “Contributing area” was considered to be visibly eroding areas near the CFT, and
slopes likely to erode into the CFT, and usually consisted of a narrow strip to each side.
43
“Hydrologic area” was basically the watershed of the CFT. Short distances were taped,
hip chained, or measured with a pole to the nearest 0.5 meter. Long or inaccessible
distances, typically the hydrologic area, were paced or estimated. Since many areas of
concentrated flow visible during high rainfall events do not have enough energy to cut
through the litter layer, the active CFTs recorded were only a small fraction of potential
contributors to runoff in the SMZ during actual rain events, and a much smaller number
than observed at the beginning of the study.
Data Analysis
Runoff frequency statistics were analyzed by storms, cups, seasons and years for
storms greater than 13 mm in depth, but the entire dataset, including smaller storms, was
used for regression analysis . Descriptive statistics (after log-x transformations ) and
regression analyses were done in Microsoft Excel ™. One storm was defined as having a
minimum 24 hour dry interval during the dormant season (Nov. 1 - April 30) and 12
hours during the growing season (May 1 - Oct 31). In cases of multiple storms between
cup checks, the data were attributed to the storm of greatest total amount, which nearly
always coincided with greater intensities. Storm durations and mean intensities were
calculated using the last or second to last tip (Onset Computer Corporation 2001). The
second last tip was used if there was a very long interval between the second last and last
tips, as compared to the preceding tips. The RUSLE “R” factor was adjusted upwards as
suggested by McGregor et al. (1995), and practiced by Ward and Jackson. (2004), for
other Piedmont study sites.
44
Table 2.1. Description of surface runoff study sites, Greene County, in Georgia Piedmont
Upland SMZ Str1 Upland SMZ Str1 Upland SMZ Str1
Area (ha) or length (km) 82 8 3.3 27 4 1.1 30 6 1.6Elevation Range2 (m) 140 -197 139 -174 145 -179 141-158 140 -179 140 -165Median Elevation2 169 152 162 149 161 148Slope Range2 (%) 0 - 20 0 - 19 0 - 23Median Slope2 (%) 6.7 4.5 1.8 6.7 8.5 2.9 8.2 6.6 3.4Stream Orders (field mapped) 0-2 0-1 0-2Perennial Stream Density (km/km2) 2.3 1.2 2.8Stand Type Pine Plt. Hdwd Pine Plt. Hdwd Nat.P/H Nat.H/PYear Established 2003 1955* 2003 1955* 1933* 1933*1Str = Stream2Elevation and slopes derived from USGS 1979 DEM, downloaded from Ga. GIS Clearinghouse* Estimated
Table 2.3. Description of cover on overland flow study sites in Georgia Piedmont
Lewis VanirSMZ1 SMZ1 Upland2 SMZ1
End of Growing Season 2 (2004) 2 (2004) (2004) (2004)Age (years) 50 Est 50 Est 72 Est 72 EstCanopy Shading (%) 88 87 90 97Total BA (sq m/ha) 23.9 24.2 24.0 25.3 Pine Basal Area 2.4 0.9 12.5 7.9 Hardwood BA 21.5 23.3 11.6 17.4Midstory (stems/ha) 2500 1500 2600 2800Ground Cover (%) Bare 0 2 10 Litter 73 87 81 Plant 20 6 3 Woody Debris 6 5 5 Gravel/Rock 0 0 1Tamped Litter Dep.(mm) 9 10 3
End of Growing Season 1 (2003) 2 (2004) 1 (2003) 2 (2004)Age (years) 0 1 0 1Max.Herb. Height (cm) 155 175 170 175Planted Pine Hgt.(cm) 150 1301SMZ conditions in first growing season assumed to be similar to second.2Cover and litter depth for upland area of Watson Springs are shown in the results section.
1 Sand, silt and clay reaching stream Sand piles near creek2 Silt and clay reaching stream Staining of leaves3 Clay reaching stream in flow of water Other visible scoured channel4 Sediments filtered out in SMZ Scoured channel ends in SMZ
before reaching stream1 Adapted from Rivenbark and Jackson (2004)
53
Figure 2.1. Location of three overland flow study sites, Greene County, in Georgia Piedmont
54
Figure 2.2 Soil map of Lewis study site, Greene County in Georgia Piedmont.
55
Figure 2.3. Soil map of Vanir study site, Greene County in Georgia Piedmont.
56
Figure 2.4. Soil map of Watson Springs study site, Greene County in Georgia Piedmont.
57
Figure 2.5. Comparison of historic rain at Watkinsville UGA Plant Farm1
1961-2002 (42 years) to surface runoff study period2,3, Greene County in Georgia Piedmont.1Watkinsville data from Ga. State Climatology Office2Missing study data supplemented by USGS data from Penfield (#02218300)3Average of study site rain gauges.
0
100
200
300
400
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Rai
n (m
m)
Watkinsville High
Watkinsville Low
Watkinsville Mean
Study 2003
Study 2004
Month
58
Figure 2.6. Diagram of typical runoff cup. a) Top view1 and b) Side view2
1Plastic washer or milk jug lid lessens shingle damage.2On steep slope, cup should be normal to slope, not vertical.
Washer
Roof tack
woodenstake
groundline
A
B
Shingle
runoffcup
59
Figure 2.7. Map of cover plot lines and grab sampling points for Lewis study site, Greene County, in Georgia Piedmont.
60
Figure 2.8. Map of cover plot lines and grab sampling points for Vanir study site, Greene County in Georgia Piedmont.
61
Figure 2.9. Map of cover plot lines for Watson Springs study site, Greene Countyin Georgia Piedmont.
62
CHAPTER 3
RESULTS AND DISCUSSION
Rainfall
Rainfall among the plantation sites was similar in the winter, due to widespread
convective patterns with long durations. It was more variable in the summer in intensity,
due to more localized thunderstorm activity (Figure 3.1). During the study period 2000
mm (79 inches) total fell at Lewis and 1728 mm (68 inches) total at Vanir, with high
variations between storms (Tables 2.4, 2.5). Comparison of intensities between sites by
seasons showed no statistical differences. However, combined data from both sites for 15
minute intensities showed a statistical difference between winter (8.1 mm/hr) and
summer (19.3 mm/hr) (Table 3.1). This would imply that for surface runoff on these sites,
Hortonian flow should be more important in the summer.
Surface Runoff Characteristics
Surface runoff frequency was highly variable within and between sites, and
between storms. Lewis had the highest fraction of cup response with a mean of 14.3%
and a standard deviation of 3.3% (Table 3.2, Figure 3.2). Only 2.1% of the storms
generated no response whatsoever, and the highest single-event response rate was over
77%. Nearly all cups (96.7%) collected runoff at least once during the study. Vanir
showed less surface runoff activity. The mean response was 6.4% with a standard
deviation of 3.0%, and a high of 39.1%. About 2% of the storms generated no response
63
and about 19% of the cups never received runoff. The reference site also had very low
runoff frequency. The mean was 2.2% with a standard deviation of 7.2% and a high of
38.0% (Table 3.3, Figure 3.2). Nearly 21% of the storms generated no runoff and over
35% of the cups never had runoff. The reference data were collected over a shorter period
but included four tropical storms. Frequency of runoff at Lewis differed statistically from
Vanir for the dormant seasons and the combined total, and from the reference site in all
categories. Vanir and the reference site were statistically similar. The combined
plantations differed statistically from the reference site for year two and combined totals.
All of the above was done by z and t-tests at alpha equal to 0.05.
Both plantations showed a drop in mean frequency from the first to the second
year, but these were not statistically significant. The Lewis median moved from 22.5% to
11.7% and Vanir from 10.1% to 5.9% (Table 3.2), for storms above 13 mm. The second
year cumulative distribution plot for the combined plantations much more closely
resembled the reference site than the first year (Figure 3.3). This decrease in runoff over
time, agreed with all similar southeastern studies. Seasonal differences of runoff
frequency were variable. The logging debris plus beds created hydraulic roughness and
depressional storage, which appeared to slow runoff and aid infiltration.
Frequency maps and field observations showed high variation in spacing and
frequency within and between storms on each site, and between sites, (Figures 3.4 - 3.10)
pointing to localized and microsite factors. More runoff was tallied in areas of
concentrated flow and variable source areas, especially at Lewis. Some of these high
frequency areas coincided with old gullies or the Chewacla soil series at Lewis (Figure
2.2), and old gullies at the other sites, but others did not.
64
Volumes of runoff were categorized by quarter runoff cups and exhibited a
bimodal distribution (Figure 3.11). Readings below ¾ cup were reasonably accurate, but
readings of full cups were only a minimum since an unknown quantity overflowed. Cups
receiving only small volumes in one storm might be full in the next storm. During large
storms, sheet flow could be 10 cm (four inches) deep. Some cups that usually had high
volumes were not located in obviously concave topography or microsites. Old stable
gullies, rills, and surface depressions leading into the SMZ were observed moving water
in large storms, but later showed no disturbance of the litter layer. So while the extremes
of sheet flow and concentrated flow were separable, the middle ground was ambiguous.
Concentrated flows were sometimes observed inundating cups during grab sampling of
large storms. Even following routine storms, fresh mud was sometimes seen on top of the
shingle cup roofs.
Possible sources of variation were small storms, interpretation of risers, and small
sample size. Smaller storms were not sampled but sometimes generated runoff, and this
accumulated in the cup until the next check, less evaporation. Cups that were displaced
from the bottom of the hole (risers) were assumed to be evidence of high water tables or
variable source areas in low areas and subsurface lateral flow (or interflow) in better
drained areas.
The greater frequency (and volume) of runoff at Lewis compared to Vanir was
attributed to three factors: finer soil textures at the runoff cups, lower landscape positions,
and slower regrowth of vegetation. Wohlgemuth et al. (2001) suggested that stream
density provides an indicator of soil storage and drainage. This may have played a role in
counterintuitive results in his runoff study on two otherwise well-matched sites. In the
65
current study, the perennial stream density at Lewis was nearly double that of Vanir
(Table 2.1).
Topographic analysis using USGS maps was foiled by coarseness of the 1:24000
scale. Extensive field experience shows that forested zero through second order streams
are sometimes missed in USGS mapping or displaced in location (Figures 3.12, 3.13)
(Hansen 2001). While mapping CFTs it became apparent that minor slope breaks on the
ground were not always indicated on the contour maps.
The low frequencies at the reference site were attributed to the evapotranspiration
pull of a mature forest and partial coverage of the site with a deep litter layer and
relatively high organic matter in the O and A horizons over most of the surface.
Conventional wisdom would hold that a mature undisturbed forest would generate runoff
only in large rain events; however some runoff was tallied even in moderate events below
15 mm (0.6 in), following a litter-disturbing prescribed burn on this tract. The frequencies
of runoff in the current study at all sites were lower than the other two frequency studies
in the tropics in mature forests (Elsenbeer and Vertessy 2000; Godsey et al. 2004).
Rainfall Runoff Relations
Since rainfall is the most important driver of runoff, regression analysis was done
on different rainfall components. Independent variables were storm depth, various
maximum intensities plus mean intensity, antecedent moisture, and storm duration,
compared with the dependent variable of runoff frequency. The “R” factor (rainfall
66
erosivity) from the Revised Universal Soil Loss Equation and the SCS Curve Number
method were also evaluated. The “R” factor was corrected by 28% for this section of the
South, as recommended by McGregor et al. (1995).
Storm depth, all maximum runoff intensities evaluated, and the R factor were
significant predictors of the response of runoff cups, with p- factors less than 0.001. The
best correlations were for storm depth, maximum storm intensities greater than six hours,
and for the R factor (Tables 3.4 – 3.6, Figures 3.14 - 3.22). The single best overall
predictor was the R factor with r2 values of 0.56 - 0.75. Storm depth had an r2 of 0.46 -
0.66, 24 hour maximum intensity 0.47 - 0.64, and six hour intensity 0.49 - 0.59. The
Lewis tract had the most runoff and best correlations. Most rain variables at both
plantations showed a decreasing effect from the first to the second years, although
differences were not always statistically significant. The reference site had a slightly
different pattern, with best fit from RUSLE “R”, but best intensities for six-hour and two-
hour intervals, and only minor differences between maximum intensities for any time
interval.
Fits of the regressions for all sites deteriorated from longer duration maximum
intensities to shorter intensities, (the 15 minute r2 = 0.27- 0.49) but these shorter duration
intensities still had predictive power, as shown by a low p-value (Table 3.4-3.6). Mean
intensity (depth/duration), storm duration, 5 or 10 day prior rainfall, and the SCS curve
number had weak or no correlations. Thirty day prior rainfall had a p-value below 0.05 at
the two plantations but the r2 was less than 0.15. This low p-value but low correlation
paralleled the experience of Wischmeier and Smith (1958), which caused them to leave
67
antecedent rainfall out of USLE. The SCS curve number greatly underpredicted runoff
frequency but improved after deletion of its threshold (initial abstraction).
It was inferred that Lewis responded more to total rainfall due to more variable
source areas caused by flatter topography near the SMZs, producing saturation overland
flow. The other two sites had runoff in higher intensity events, as the Hortonian overland
flow model suggests. The lower coefficients of determination (r2) for shorter time
maximum intensities were likely due to higher seasonal variability, noted above in the
rainfall section. Unfortunately, good seasonal data for the first summer were only
available from Vanir, which did not have much runoff at all. The reference site, with
more evotranspiration demand, less compaction and perturbation, and presumably more
macropores, could infiltrate most rainfall. Consequently, it had runoff only in the higher
intensity events, and was less influenced by antecedent soil moisture. RUSLE “R” is
based on a combination of kinetic energy of the storm, 30 minute maximum rainfall
intensity, and indirectly, rainfall duration, and therefore blends factors. Since the 30-
minute maximum intensities were less highly correlated than most other intensities, and
rain duration had no relationship in this study, the energy portion of the equation
presumably made up for these deficiencies.
Ground Cover
Despite the application of herbicides eight months before planting, vegetation
made a rapid recovery throughout the first spring and summer, aided by plentiful rainfall
(Figure 2.5). Bare ground at Lewis was 31% after the first growing season and 27% after
the second growing season. Bare ground at Vanir was 27% after the first growing season
68
and 11% after the second growing season (Table 3.7, Figure 3-23). The reference site had
29% bare ground the second year of the study, due to a prescribed burn over part of the
area but still averaged 5 mm litter depth (Figure 3.10). The plantation litter depth layer
increased from 2 to 4 mm from the first to second years. The tallest brush in the
plantations was 178 cm and planted pines were 138 cm high at the end of the second
growing season (Table 2.3).
Revegetation has many effects on water quality. On the plus side, the canopy
breaks the impact of rainfall, and returns some directly to the atmosphere by interception.
Vegetation creates evapotranspiration demand, and therefore drier soil between storms in
the summer, causing more infiltration. Stems create hydraulic roughness, which slow
down runoff and leaf drop contributes to the litter layer. Litter breaks the force of rain
drops and acts as a sponge, slowing water movement by absorption and adsorbtion,
allowing more infiltration and creating hydraulic roughness. Organic matter has a high
surface area similar to fine clays (Brady and Weil 1999). On the minus side, runoff
carrying leaf litter and other plant byproducts can export nutrients from the site,
particularly water soluble nitrates. Also, herbaceous competition for on-site resources
slows pine growth, so foresters like to “capture the site” by intensive site preparation and
use of herbicides, baring more soil longer.
Although not formally studied, revegetation seemed to be the greatest factor in
reducing runoff over the study period for both sheet and concentrated flow. This was
particularly noticeable in the variable source areas at Lewis, where logging ruts were
covered by a dense stand of sedges and other plants by the end of the first summer. At the
69
reference site, of 23 active CFTs in early August 2004, 14 had stabilized nine months
later, after leaf drop and spring growth.
Water Chemistry
Grab samples of plantation runoff were taken during six large storms and
analyzed for soluble nitrates, soluble phosphates, and total suspended solids. Differences
in water chemistry between plantations were minor, so the results were combined by
year. Soluble nitrates decreased from the first to the second year (mean 2.1 versus 0.8
mg/L and median 3.4 versus 0.0 mg/L), but the number of samples was small in the first
year (Figure 3.24). The phosphate mean was 0.21 mg/L the first year and 0.12 the second
year, and the median 0.23 the first year, and 0.17 mg/L the second year. TSS data were
highly variable between samples and storms; with up to a three order of magnitude range
(Figure 3.25). The first and second year means were 54 and 36 mg/L, and the medians
were 42 and 29 mg/L. Two samples were greater than 1000 mg/L, with the maximum
1436 mg/L, both from active concentrated flow tracks. TSS data from fixed collectors
(mean = 66 and median = 80 mg/L) were only obtained during the second year and were
slightly higher than the grab samples, although not statistically different. Collector data
were cumulative throughout each storm, and between storms. Differences between years
were not statistically significant for any of the water chemistry data in the same category.
EPA recommended targets for Southern Piedmont perennial streams at baseflow
are 0.17 mg/L for nitrates, 0.01 mg/L for phosphates, and 6 NTU for turbidity (US
Environmental Protection Agency 2000), which is equivalent to 6 mg/L TSS, in this area
(Barnes 1998). Observed stormflow nutrients and sediment were 9-18 times these
70
recommendations for perennial streams the first year, and 5-11 times these
recommendations the second year. However, observed values were in the same range as
other southeastern regeneration studies of ephemeral streams and runoff plots (Figures
in nutrients over time, typically one to three years.
Higher sediment concentrations from the collectors seemed somewhat
counterintuitive because sheet flow would have less transport power and velocity than the
grab samples of more concentrated flow. However these collectors were constructed in
places assumed conducive to sheet runoff in terms of slope and bare soil, and measured
quasi-cumulative sediment from storms, since sediment did not always wash through. By
contrast, the grab samples represented one instant in time, usually several hours into the
storm and long past first flush. For comparison, a current study of 42 third through
seventh order streams in the Georgia Piedmont showed base flow TSS with a mean of 6
and a median of 5 mg/l, from 454 samples (G. Denise Carroll, unpublished data). The
larger perennial streams in that study have a lot more transport power than the tiny
ephemeral gullies and flow tracks in this study, yet have low TSS readings.
Collection of water quality data presented several problems. Grab sampling
shallow tracks and puddles usually stirred up some sediment, in spite of best efforts.
Available locations with concentrated flow for grab sampling decreased over the course
of the study, and by the end could only be found in the most active rills and gullies. The
metal overland flow collectors were designed for agricultural research and seemed overly
elaborate in terms of capacity, expense, installation, and servicing requirements for a
descriptive forestry study generating only small amounts of runoff. An attempt to
71
establish pitfall traps in gullies failed, due to the buckets floating and the severity of the
environment. However, at least one researcher has reported success with pitfalls in
streams by use of a securely fastened outside collar and wedging the inside collector
(Sutherland et al. 2002).
Concentrated Flow Tracks (CFTs)
Active concentrated flow tracks (as shown by a scoured path through the litter)
decreased in power and number during the study period. In a large rainstorm during the
first spring, 84 areas of concentrated flow were observed at Lewis (or an average of
one per 44 m of SMZ boundary), but few of these scoured to mineral soil. Five scoured
CFTs were tallied at the beginning of the second summer and there were only three left
when measured towards the end of the second growing season (Figure 3.29). At Vanir,
seven active CFTs were observed during a large storm two months post-planting (March
2003), two were measured in the spring of 2004, and all were stabilized by the late
summer. In contrast, the reference site had 23. This relatively high number was attributed
to steeper slopes (Table 2.1) and reactivation of old gullies due to burning. Of the 23, 18
were old agricultural gullies, four were attributed to converging topography, and one was
mixed. The burned area contained 16 of the 23 CFTs, including 13 of the old gullies
(Figure 3.30). Most of these stabilized, following leaf fall and spring greenup, leaving
only six active by May 2005 (Figure 3.31), prior to a planned timber cut. Even these
remaining active ones were observed to have more clogged flowpaths than previously,
decreasing speed and cutting power of runoff. However, the trend in CFTs was not
uniformly toward stabilization. Over the study, some were observed to open up or
72
lengthen at least temporarily after big storms, showing the dynamic between vegetative
regrowth and the erosive force of water. Also, even stabilized CFTs and old gullies could
move significant quantities of water in large storms without scouring the leaf litter down
to soil.
Statistical differences among attributes of the CFTs were inconclusive due to high
variation and small numbers involved. Generally the CFTs that carried sediment to the
creek were shorter, deeper, and drained a larger area than the ones that did not (Table
3.8). Also, most CFTs had more bare ground on the bottom and sidewalls than on the
actively eroding slope and nearby areas. Comparisons of the current study to the previous
study of 30 forest regeneration sites in the Georgia Piedmont (Rivenbark and Jackson
2004) were likewise inconclusive .
All three of the active CFTs at Lewis connected into previous agricultural gullies
for at least part of their length. Two were very long, at 106 and 146 meters (348 and 479
ft). These drained either roads or decks at ridge or shoulder landscape positions, and
traveled all the way down the 10% backslope to the toeslope and floodplain, funneling
into previous gullies on the lower part of the slope (Figure 3.29). One was stopped by the
SMZ and the other fed directly into a deep gullied tributary. Even the latter had some
sediment filtered by a fortuitous pile of logging slash at the edge of the SMZ. The first
was not a complete success, although it normally did not tie into the stream. The
sediment which it deposited on the floodplain would likely remobilize during future big
storms. The third 3.8 meter deep gully was at the downhill end of a long series of old
gullies (Figure 3.29). These were partly stabilized on the bottom but had the potential to
contribute sediment during major events, especially from the sidewalls, and probably
73
were major contributors early in the study. However, it also should be noted that nearly
all flow tracks on both plantations were stabilized after two seasons of regrowth without
herbicides.
The 70 year old reference forest would not normally be considered a major
sediment source. However, it became a temporary contributor, following a burn, and
given old agricultural gullies. When the runoff cups on site were first established, the
streambed was relatively clean and stable. Following the burn, fresh sand deposits were
observed all throughout the stream system in the study area, although it is unclear how
much was contributed from firelanes, sheet flow from the burn area, highway
embankment runoff, and CFTs (Figure 3.30). Some deposits were upstream of all
firelanes and active CFTs.
Implications and Considerations for Management
Of the many factors influencing runoff, the land manager typically only has some
short run control of vegetative cover, management practices, and land use. Even mature
forests generate some runoff and sediment, especially in connection with logging roads,
skid trails, fire lanes, and during the cutting and regeneration window. Walking around
on a tract with all-weather access during a heavy rainstorm is a good way to raise
awareness of runoff, where it happens, and how the road drainage system is functioning.
In the Piedmont, old gullies are ubiquitous (Trimble 1974) and may be reactivated
by any land disturbing activity (e.g. Hewlett et al. 1984). These activities may also create
independent concentrated flow tracks, some of which may create new gullies or tie into
old ones. Flow tracks, ruts and small depressions are often present and can also channel
74
water. People working on the land should be aware of impacts to sensitive areas. These
include wetlands and wet areas, deep gullies, active gullies, gully “nests”, toeslope areas,
and areas of convergent topography (vallies). Compaction should be minimized in
sensitive areas by proper road and deck placement. The question of “How do we get the
water off the road?” needs to be followed by “Where is the drain water going?” In some
cases, it might be necessary to flatten and/or revegetate a place at the end of wing ditches
or below decks on steep hillsides to facilitate infiltration and dropping of sediment load.
Examples of management for water quality for gullies might include leaving at
least a thin “picket fence” border of brush or trees on the sides, and especially the heads,
and minimizing roads, compaction, and intensive site preparation in the immediate area.
Herbicides should be minimized in this low impact area, since revegetation is so
important in preventing runoff. If logging debris were being piled, it could be pushed into
gullies that do not penetrate to groundwater, to help slow runoff. However, debris should
be moved only short distances to the nearest small or medium sized gully, to minimize
tract compaction and risk of washing debris downstream during floods in large gullies,
with possible bank and structural damage. Slash should be pushed only horizontally or
uphill, to avoid convergent water flow paths. Also it should be pushed into gully sides,
with minimal disturbance to gully cover, leaving the gully head undisturbed. The areas
freed up for planting would help compensate for the less intense management in the
sensitive areas, and improve management access. Site preparation rakes and blades
should be run a little above ground to minimize moving dirt and cover, even if the results
look sloppy due to small limbs dropping on the ground.
75
In Georgia, along with some other southeastern states, recommended SMZ widths
are based on slope gradient (Table 2.8). While these guidelines are reasonably effective
for sheet flow, areas of concentrated flow due to old gullies or converging terrain
(vallies) can exceed even the maximum 22 meter (70 ft) non-trout stream buffer width, as
shown by long travel distances on this study. Toeslopes, wet areas, and concentrated flow
tracks route a disproportionate share of water into the SMZ. A more site-specific
prescription might widen the SMZ to include sensitive areas and narrow it in less
problematic places. On the plantations studied, a short uphill extension of the SMZ while
delineating the timber sale boundary would have protected places where deep gullies
intersected ground water and small seeps near the SMZ, since these sensitive areas were
close to the delineated boundary. Some management activities on sensitive tracts could
possibly be timed to take advantage of green up or leaf drop for site stabilization.
76
Table 3.1. Comparison of rain by seasons for combined plantations,Greene County, Ga., in Georgia Piedmont, 2/2003 - 9/2004.
Mean Median Std DevRain Depth (mm) 10.6 12.2 3.3 Winter1 11.4 13.2 2.9 Summer1 10.7 10.7 3.224 Hour Maximum Intensity (mm/hr) 0.4 0.4 3.0 Winter 0.4 0.5 2.8 Summer 0.4 0.4 3.315 Minute Maximum Intensity (mm/hr) 11.8 11.2 2.8 Winter (a) 8.5 8.1 2.5 Summer (b) 15.5 19.3 2.81Winter = November - April; Summer = May - October(a) (b) Means and medians are statistically different between seasons at alpha= 0.05
77
Table 3.2. Descriptive statistics of response cups in plantations by seasons and years, for storms above 13 mm threshold, in Greene County, GeorgiaPiedmont overland flow study.
Combined Combined Combined CombinedStatistic1 Dormant2 Growing Year 1a Year 2 Total
Runoff Frequency per Storm (%) Mean3,4 17.0 12.7 20.3 10.9 14.3 Standard Deviation 1.9 4.0 3.0 3.3 3.3 Median 20.5 16.0 22.5 11.7 17.4 Maximum 52.5 77.1 77.1 59.3 77.1Cups Without Runoff in any Storm (%) 22.0 7.3 8.9 18.7 3.3Number of Storms Studied 17.0 31.0 21.0 27.0 48.0
Runoff Frequency per Storm (%) Mean 6.2 6.6 8.7 4.7 6.4 Standard Deviation 2.5 3.5 2.6 3.3 3.0 Median 6.0 7.8 10.1 5.9 7.6 Maximum 39.1 30.9 39.1 30.3 39.1Cups Without Runoff in any Storm (%) 36.2 24.6 30.4 37.7 18.8Number of Storms Studied 19.0 26.0 23.0 22.0 45.01Data were log x transformed for statistics.2Dormant = Nov - April; Growing = May - Oct; both years included.aYear 1 = Feb - Oct 2003 ; Year 2 = Nov 2003 - Sep 2004 for plantations. 3Means are statistically equivalent within sites.4Lewis and Vanir means statistically differ for the same column for dormant and total.
---------------------------Lewis Study Site-----------------------------
---------------------------Vanir Study Site------------------------------
78
Table 3.3. Comparison of response cups in plantations to reference site by seasons and years,for storms above 13 mm threshold, in Greene County, Georgia Piedmont overland flow study
Combined CombinedStatistic1 Dormant2 Growing Year 1a Year 2 Total Dormant Growing Year 2b
Runoff Frequency per Storm (%) Mean3,4 10.0 9.5 13.0 7.4 9.7 1.4 2.5 2.2 Standard Deviation 2.6 3.9 3.0 3.5 3.4 5.3 7.9 7.2 Median 10.7 11.8 16.2 9.1 11.6 1.3 2.5 1.8 Maximum 52.5 77.1 77.1 59.3 77.1 8.2 38.0 38.0Cups Without Runoff in any Storm (%) 27.1 13.5 16.7 25.5 8.9 88.6 35.4 35.4Number of Storms Studied 36.0 57.0 44.0 49.0 93.0 5.0 19.0 24.01Data were log x transformed for statistics.2Dormant = Nov - April; Growing = May - Oct; both years included for plantations, only second year for reference.aYear 1 = Feb - Oct 2003 ; Year 2 = Nov 2003 - Sep 2004 for plantations. bYear 2 = Feb-Sep 2004 for reference.3Means are statistically equivalent within sites at alpha= 0.05.4Means statistically differ between sites for total at alpha= 0.05.
Table 3.4. Investigations of surface runoff frequency (%) as explained by precipitation metrics, at Lewis site, in Greene County, Georgia Piedmont, all storms tallied.
Independent Variable P-value r2 Best Fit Equation1 No.of Res. Res.Storms Outliers Bias
Storm Depth - Both yrs (mm) * 0.66 0.5273 x + 2.976 (ab) 65 0 N1st Yr * 0.78 0.6312 x + 5.630 (a) 28 0 N2nd Yr * 0.69 0.4096 x + 1.687 (b) 37 1 ?
RUSLE "R" (English) * 0.56 0.6592 x + 9.878 44 0 NMaximum Intensities (mm/hr) 44 24 Hr * 0.64 11.26 x + 3.900 0 N 12 Hr * 0.60 6.386 x + 3.755 0 N 6 Hr * 0.59 4.092 x + 2.454 1 N 2 Hr * 0.50 1.989 x + 1.592 0 N 1 Hr * 0.32 0.9543 x + 5.294 0 N 30 Min * 0.24 0.4584 x + 7.774 0 N 15 Min * 0.27 0.3200 x + 7.367 0 N Mean Int. 0.663 44Duration (hrs) * 0.24 0.5404 x + 8.568 44 1 YPrior Rainfall (cum mm) 30 Day 0.005 0.13 0.0862 x + 7.376 58 1 ? 10 Day 0.072 - CI for slope included 0 62 1 Y 5 Day 0.154 - CI for slope included 0 63 1 YSCS Curve # Runoff Depth (in) Unusable 51 0 YSCS Curve # with no threshold (in) Unusable 51 0 Y* P-values less than 0.001.1Slopes of multiple equations for the same independent variable with the same letter are statistically equivalent at alpha= 0.05.
80
Table 3.5. Investigations of runoff frequency (%) as explained by precipitation metrics, at Vanir site, Greene County, Georgia Piedmont, all storms tallied.
Independent Variable P-value r2 Best Fit Equation1 No.of Res. Res.StormsOutliersBias
Storm Depth - both years (mm) * 0.51 0.2221 x - 1.424 (a) 58 1 N1st year * 0.53 0.2942 x +2.175 (a) 28 1 N2nd year * 0.78 0.1874 x - 0.1675 (a) 30 0 N
RUSLE "R" - both years (English) * 0.54 0.5255 x + 4.400 (a) 49 1 N1st year * 0.54 0.5789 x + 6.185 (a) 26 0 N2nd year * 0.67 0.4445 x + 2.459 (a) 23 0 ?
Intensities (mm/hr) 24 Hr - both years * 0.51 5.508 x + 2.1639 (a) 49 1 N
1st year * 0.57 7.968 x +2.225 (a) 26 1 ?2nd year * 0.78 4.424 x + 0.2588 (a) 23 0 N
12 Hr - both years (a) * 0.49 3.310 x + 1.910 (a) 49 1 N1st year (b) * 0.67 6.392 x - 0.8970 (b) 26 0 N2nd year (a) * 0.81 2.566 x + 0.0539 (a) 23 0 N
6 Hr - both years (c,d) * 0.49 2.341 x + 0.7884 (a,b) 49 1 N1st year (c) * 0.63 4.078 x - 2.134 (a) 26 0 N
2nd year (d) * 0.74 1.764 x - 0.3872 (b) 23 0 N 2 Hr - both years * 0.51 1.300 x - 0.7940 (a) 49 0 N
1st year * 0.64 1.805 x - 2.554 (a) 26 0 N2nd year * 0.52 0.8893 x - 0.3510 (a) 23 0 N
1 Hr * 0.48 0.8028 x - 0.2420 49 0 N 30 Min * 0.41 0.4802 x - 0.8510 49 0 N 15 Min * 0.28 0.2898 x + 0.1.560 49 0 N Mean Int. 0.592 - CI for slope included 0 49 1 YDuration (hrs) 0.094 - CI for slope included 0 49 1 YPrior Rainfall (cum mm) 30 Day 0.010 0.13 0.0005 x + 0.0312 52 0 ? 10 Day 0.695 - CI for slope included 0 54 0 Y 5 Day 0.174 - CI for slope included 0 55 1 ?SCS Curve # Runoff Depth (in) Unusable 55 YSCS Curve # with no threshold (in) * 0.30 16.12 x + 4.831 55 0 Y* P-value less than 0.001.1Slopes of multiple equations for the same independent variable with the same letter are statistically equivalent at alpha= 0.05.
81
Table 3.6. Investigations of surface runoff frequency(%) as explained by precipitation metrics, at Watson Springs site,Greene County, in Georgia Piedmont, all storms tallied.
Independent Variable P-value r2 Best Fit Equation No.of Res. Res.StormsOutliers Bias
Storm Depth (mm) * 0.46 0.2507 x - 1.224 27 1 NRUSLE "R" (English) * 0.75 0.4478 x - 1.226 24 0 NIntensities (mm/hr) * 24 24 Hr * 0.47 6.172 x - 1.205 1 N 12 Hr * 0.50 3.459 x - 1.532 1 N 6 Hr * 0.59 2.347 x - 3.621 1 N 2 Hr * 0.58 1.186 x - 5.911 0 N 1 Hr * 0.50 0.6636 x - 5.506 0 N 30 Min * 0.42 0.3516 x - 3.930 0 N 15 Min * 0.49 0.2555 x - 4.324 0 N Mean Int. 0.552 24 0 YDuration (hrs) 0.888 24 0 YPrior Rainfall (cum mm) 30 Day 0.075 CI for slope included 0 27 0 N 10 Day 0.256 27 0 N 5 Day 0.843 27 0 YSCS Curve # Runoff Depth (in) - - Unusable 27 - YSCS Curve # with no threshold (in) * 0.39 20.64 x + 3.094 27 1 ?* P-value less than 0.001.
82
Table 3.7 . Comparison of cover and litter depth over time between surface runoff study sites, Greene County, in Georgia Piedmont
Cover (%) Bare 27 11 22 29 Gravel 0 1 1 0 Litter 26 53 34 64 Plant 39 26 35 3 Wood 7 8 8 3Standard Error 3 3 2 5Litter Layer (mm) 4 5 4 51 Plantations were clearcut, herbicided, spot piled, ripped/bedded, andhand planted prior to study.2Ca. 70 year old pine-hardwood reference stand was partly burned prior tosecond growing season.See figure3.23 for graphic illustration.
----------End of first growing season (year 1)----------
-------End of second growing season (year 2)--------
83
Table 3.8. Summary of active1 concentrated flow tracks for combined Lewis and Watson Springs sites, Greene County, in Georgia Piedmont2,3
Infiltration5
Sand Silt ClayNumber 19 0 2 5 Old Ag Gullies 15.5 1 4.5 Topographic CFTs 3.5 1 0.5Length (m) 29 71 44Depth (m) 0.9 0.5 0.6Land Slope (%) near CFT 13 14 14Land Slope (%) of Hyd. Cont. Area (CA) 14 8 10Bare Ground (BG) Nearby (%) 15 12 24Bare Ground Sidewalls of CFT (%) 58 59 34Bare Ground Bottom of CFT (%) 39 63 42SMZ incursion (m) 12.3 10.2 5.6CFT Area above SMZ (m2) 15 55 22Hydrologic Contributing Area (ha) 0.4 0.8 0.2BG*CA (ha) 3 7 6BG*CA*Slope (ha) 36 57 63BG*CA (ac) 8 12 16BG*CA*Slope (ac) 89 96 1571Active tracks had a scour which intruded into the Streamside Management Zone.2All data are medians, except counts.3Data from Lewis was two growing seasons after planting and Watson Springs one growing season after burn.4 Breakthroughs had active scours fully penetrating the SMZ.5 Scour disappeared in the SMZ before reaching the stream.
----------Type Breakthrough4---------
84
Figure 3.1. Comparison of variation in storm depths between plantations1 and storms by seasons, Greene County, in Georgia Piedmont.1 The plantations are only 13 km (eight miles) apart.
Summer Storm Depths
0
20
40
60
80
100
120
140
Storms
Prec
ipita
tion
(mm
)
Lewis VanirMay - Oct 2003 May - Sept 2004
Winter Storm Depths
020406080
100120140
Storms
Prec
ipita
tion
(mm
)
Lewis VanirFeb - April 2003
Nov 2003- April 2004
85
Figure 3.2. Mean response cups in plantations by years, for storms above 13 mm threshold, in Greene County, Ga. Piedmont surface runoff study1Top row is number of storms > 13 mm threshold for site below.2Second row is number of cups per site.*Whiskers are standard error#Frequencies were log transformed for statistics
0
5
10
15
20
25
Year 1 Year 2
Ron
off F
requ
ency
(%)
LewisVanirComb. Plt.Reference
2221 23 27 24No. Storms1
123 69 79No. Cups2
49
192
44
*
#
86
Figure 3.3. Cumulative distribution of surface runoff frequency for storms >13 mmcombined plantations1plus reference site2, Greene County in Georgia Piedmont.1Year one was the period for nine months post-planting.Year two was the following 11 months.2 The reference site data were for most of year two.
0
10
20
30
40
50
60
70
80
90
100
5 15 25 35 45 55 65 80 85 95
Runoff Frequency (%)
Per
Cen
t Les
s Tha
n
ReferencePlt - Yr 2Plt - Yr 1
87
Figure 3.4. Surface runoff frequency for entire study period1by cups, Lewis tract, Greene County, in Georgia Piedmont. 1Post-planting through most of second growing season.
88
Figure 3.5. Surface runoff frequency for “year one”1 by cups, Lewis tract, in Greene
County, Georgia Piedmont. 1 Post-planting through end of first growing season.
89
Figure 3.6. Surface runoff frequency for “year two”1 by cups, Lewis tract, Greene
County, in Georgia Piedmont. 1 End of first growing season through most of second growing season.
90
Figure 3.7. Surface runoff frequency for entire study period1 by cups, Vanir tract, Greene County, in Georgia Piedmont. 1Post-planting through most of second growing season.
91
Figure 3.8. Surface runoff frequency for “year one”1 by cups, Vanir tract,
Greene County, in Georgia Piedmont. 1Post-planting through end of first growing season.
92
Figure 3.9. Surface runoff frequency for “year two”1 by cups, Vanir tract,
Greene County, in Georgia Piedmont. 1Beginning of second dormant season through most of second growing season.
93
Figure 3.10. Surface runoff frequency for “year two”1 by cups, Watson Springs (reference) tract, Greene County, in Georgia Piedmont.
1Part of study second dormant season through most of study second growing season, in mature timber.
94
Figure 3.11. Distribution of runoff cup volumes for all three study sites combined, if surface runoff occurred, Greene County, in Georgia Piedmont.
0100200300400500600700800
0-25% 25-50% 50-75% Full orFlooded
Risers
Volume in Cups
Cou
nt ReferenceVanirLewis
95
Figure 3.12. Comparison of field mapped water features to USGS topographic
map features at Watson Springs tract, Greene County, Georgia, Greshamville 7.5’ Quadrangle.
96
Figure 3.13. Comparison of field mapped water features to USGS topographic
map features at Lewis tract, Greene County, Georgia, Greshamville 7.5’ Quadrangle.
97
Figure 3.14. Storm depth versus surface runoff frequency at Lewis tract in Georgia Piedmont, all tallied storms included a) All data b) Data separated by year1,2.1Year one is first dormant and growing season post-planting2 Year two is second dormant and most of second growing season post-planting.
All Data y = 0.5273x + 2.9757R2 = 0.66
0
10
20
30
40
50
60
70
80
0 20 40 60 80 100 120 140 160
y = 0.4096x + 1.6867R2 = 0.69
y = 0.6312x + 5.6296R2 = 0.78
0
20
40
60
80
0 20 40 60 80 100 120 140
Storm Depth (mm)
Cup
Res
pons
e (%
)
Year 2
Year 1
By Years
98
Figure 3.15. Maximum rain intensities versus surface runoff frequency at Lewis tract in Georgia Piedmont, all tallied storms included a) 24 hour and b) 6 Hour
y = 4.0922x + 2.4537R2 = 0.59
0
20
40
60
80
0 2 4 6 8 10 12 14 16 18
Maximum 6 Hr Intensity (mm/hr)
Cup
Res
pons
e (%
)
y = 11.259x + 3.8951R2 = 0.64
0
20
40
60
80
0 1 2 3 4 5 6 7
Maximum 24 Hour Intensity (mm/hr)
Cup
Res
pons
e (%
)
99
Figure 3.16. Storm depth versus surface runoff frequency at Vanir tract in Georgia Piedmont, all tallied storms included a) All data b) Data separated by year1,2.1Year one is first dormant and growing season post-planting2 Year two is second dormant and most of second growing season post-planting.
y = 0.2221x + 1.4236R2 = 0.51
0
10
20
30
40
50
0 20 40 60 80 100 120 140 160
All Data
By Years y = 0.2942x + 2.1747R2 = 0.53
y = 0.1874x - 0.1675R2 = 0.78
0
10
20
30
40
50
0 20 40 60 80 100 120 140 160
Storm Depth (mm)
Cup
Res
pons
e (%
)
Year 11
Year 2
100
Figure 3.17. Maximum 24 hour intensity versus surface runoff frequency at Vanir tract in Ga. Piedmont, all tallied storms included a) All data b) Data separated by year1,2.1Year one is first dormant and growing season post-planting2 Year two is second dormant and most of second growing season post-planting.
All Data y = 5.5078x + 2.1639R2 = 0.51
0
10
20
30
40
50
0 1 2 3 4 5 6 7
Maximum 24 Hr Intensity (mm/hr)
Cup
Res
pons
e (%
)
By Years y = 7.9684x + 2.2252R2 = 0.57
y = 4.4239x + 0.2588R2 = 0.78
0
10
20
30
40
50
0 1 2 3 4 5 6 7Maximum 24 Hr Intensity (mm/hr)
Cup
Res
pons
e (%
)
Year 1
Year 2
101
Figure 3.18. Maximum 6 hour intensity versus surface runoff frequency at Vanir tract in Georgia Piedmont, all tallied storms included a) All data b) Data separated by year1,2.1Year one is first dormant and growing season post-planting2 Year two is second dormant and most of second growing season post-planting.
Figure 3.19. Rain factors versus surface runoff at reference site in Georgia Piedmont,all storms tallied, a) Storm depth b) 24 hour maximum intensity c) 6 hour maximum intensity, for 8 of 11 months of year two of study.
y = 0.2507x - 1.2244R2 = 0.46
0
10
20
30
40
0 20 40 60 80 100 120 140 160
Storm Depth (mm)
Cup
Res
pons
e (%
)
y = 6.1721x - 1.2049R2 = 0.47
0
10
20
30
40
0 1 2 3 4 5 6
Maximum 24 Hour Intensity (mm/hr)
Cup
Res
pons
e (%
)
y = 2.3474x - 3.6213R2 = 0.59
0
10
20
30
40
0 2 4 6 8 10 12 14 16
Maximum 6 Hour Intensity (mm/hr)
Cup
Res
pons
e (%
)
103
Figure 3.20. RUSLE "R"1 versus surface runoff frequency at Lewis tract in GeorgiaPiedmont, all tallied storms included 1Storms were separated by 24 hour gaps in the winter and 12 hour gaps in the summer, not by RUSLE standard of 6 hour gaps.
y = 0.6592x + 9.8782R2 = 0.56
0
20
40
60
80
0 10 20 30 40 50 60 70 80 90
"R" Factor (SI Units)
Cup
Res
pons
e (%
)
104
Figure 3.21. RUSLE "R"1 versus surface runoff frequency at Vanir tract in GeorgiaPiedmont, all tallied storms included a) All data b) Data separated by years2.3
1Storms were separated by 24 hour gaps in the winter and 12 hour gaps in the summer, not by RUSLE standard of 6 hour gaps.2Year one is first dormant and growing season post-planting3 Year two is second dormant and most of second growing season post-planting.
By Years y = 0.5789x + 6.1852R2 = 0.54
y = 0.4445x + 2.4589R2 = 0.67
0
10
20
30
40
50
0 10 20 30 40 50 60 70 80
"R" Factor (SI Units)
Cup
Res
pons
e (%
) Year 1
Year 2
All Data y = 0.5255x + 4.4003R2 = 0.54
0
10
20
30
40
50
0 10 20 30 40 50 60 70
"R" Factor (SI Units)
Cup
Res
pons
e (%
)
105
Figure 3.22. RUSLE "R"1 versus surface runoff frequency at reference site in Georgia Piedmont, all tallied storms included, for 8 of 11 months of year two of study. 1Storms were separated by 24 hour gaps in the winter and 12 hour gaps in the summer, not by RUSLE standard of 6 hour gaps.
y = 0.4478x - 1.2261R2 = 0.75
0
10
20
30
40
50
0 10 20 30 40 50 60 70 80 90 100
"R" Factor (SI Units)
Cup
Res
pons
e (%
)
106
Figure 3.23. Cover and litter depth by sites and years2, Greene County, in GeorgiaPiedmont.1Reference site is one growing season post-burn.Litter depth is on secondary axis with different scale.2Graph represents data in table 3.7
0
20
40
60
80
100
1 2 1 2 1 2 2Site and Year
Cov
er (%
)
0
10
20
30
40
50
Litt
er D
epth
(mm
)
GravelWoodLitterPlantBareLitter Depth
Lewis Vanir Combined Plt. Reference1
107
Figure 3.24. Boxplots of combined concentrations of dissolved a) Nitrates1 and b) Phosphates grab sampled during large storms at pine plantation concentrated flow tracks, Greene County, in
Georgia Piedmont2. 1 First year nitrates capped at 6.0 mg/L due to test limitations. 2 Year one is within first year post-planting, year two is within second year post -planting. Note log scale on y-axis. Boxes are 25 percentile and 75 percentiles. Whiskers are 10 and 90 percentiles. Solid lines are medians and dashed lines within box are (means). The long dashed line is EPA baseflow standard for Southern Piedmont perennials (US EPA 2000) .
n = 17 n = 58
6.0
(2.1)3.4
1.7
1.0
(0.1)
0.0, 0.0
Year1 2
Solu
ble
Nitr
ates
Con
c (m
g/L
)
0.01
0.1
1
10
n = 33 n = 58Year1 2
Solu
ble
Phos
phat
eC
onc.
(mg/
L)
0.01
0.10
1.00
0.310.23, (0.21)0.16 (0.12)
0.28
0.17
0.10
108
Figure 3.25. Boxplots of concentrations of Total Suspended
Solids (TSS) sampled during storms at two pine plantations, Greene County, in Georgia Piedmont1.
1 Year one is within first year post-planting, year two is within second year post-planting.
The two left boxes are data from grab samples in concentrated flow tracks at the Lewis and Vanir sites.
The right box is data from overland flow collectors at the Vanir site.
Note log scale on y-axis. Boxes are 25 percentile and 75 percentiles. Whiskers are 10 and 90 percentiles. Solid lines are medians and dashed lines within box are (means).
The long dashed line is the EPA baseflow standard for Southern Piedmont perennials (US EPA 2000).
n = 33 n = 55 n = 26
Grab Grab Collector
Year
1 2 2
TSS
Con
c.(m
g/L
)
1
10
100
1000
10000
2742
89
(54)
11
29(36)
80 (66)
12580
48
109
Figure 3.26. Mean NO3 stormflow a) Concentrations and b) Loads in
southeastern US forested ephemeral streams or runoff plots, from calibration through second year post-treatmentBlank spaces are successive years or a different site in the previous study number.* = Significant difference between treatment and control for that year.No data shown as 0.001. Note log scale on y-axis. 21 = current study. See Table 1.3 for citations.Labels show most site disturbing but not every treatment. Re= Reference for Eastern perennials;Un=Uncut calibrations; B= Burn; CC= Clearcut; C = Chop; Be = Bed; Pi= Pile; Di = Disc
0.00
0.01
0.10
1.00
10.00
10Re
17 4Un
5 19 5Bu
19CC
7 8 4*C
21Be
4*Pi
* * 7*Di
8* *
Studies and Maximum Treatments
Con
c. (m
g/L)
TreatmentControl
0.00
0.01
0.10
1.00
10.00
10Re
17 4Un
5 19 5Bu
19CC
7 * 8 4C
21Be
4*Pi
* * 7*Di
8
Studies and Maximum Treatments
Load
(kg.
ha/y
r)
TreatmentControl
110
Figure 3.27. Mean PO4 stormflow a) Concentrations and b) Loads in
southeastern US forested ephemeral streams or runoff plots, from calibration through second year post-treatmentBlank spaces are successive years or a different site in the previous study number.* = Significant difference between treatment and control for that year.No data shown as 0.001. Note log scale on y-axis. 21 = current study. See Table 1.3 for citations.Labels show most site disturbing but not every treatment. Re= Reference for Eastern perennials;Un=Uncut calibrations; B= Burn; CC= Clearcut; C = Chop; Be = Bed; Pi= Pile; Di = Disc
0.00
0.05
0.10
0.15
0.20
0.25
13Re
4Un
5 19 5 7CC
19 7 4C
21Be
7Di
4*Pi
Studies and Maximum Treatments
Con
c. (m
g/L
)
TreatmentControl
0.000.010.010.020.020.030.030.040.040.05
13Re
4Un
5 19 5 7CC
19 7 4C
21Be
7Di
4*Pi
Studies and Maximum Treatments
Loa
ds (k
g/ha
/yr)
TreatmentControl
111
Figure 3.28. Mean sediment stormflow a) Concentrations and b) Loads in southeastern US forested ephemeral streams or runoff plots, from calibration through second year post-treatment* = Significant difference between treatment and control for that year.No data shown as 1. Note log scale on y-axis. See Table 1.3 for citations.Labels show most site disturbing but not every treatment. Re= Reference for Eastern perennials;Un=Uncut calibrations; B= Burn; CC= Clearcut; C = Chop; Be = Bed; Pi= Pile; Di = Disc
21 Current study
1
10
100
1000
10000
14Re
2Un
6 3 18 1 5 19 5B
12C
7 1111 618 19*
16 9 7 6 1 3C
9Be
21 20 1Pi
3* * 2 7*Di
15
Studies and Maximum Treatments
Con
c. (m
g/L
)TreatmentControl
1
10
100
1000
10000
100000
14Re
2Un
6 3 18 1 5195B
12*
7* 1111618 19*
16 9 7 6 1 3C
9Be
21 20 1Pi
3* * 2 7*Di
15
Southeastern US Forest Studies
Loa
d (k
g/ha
/yr)
TreatmentControl
112
Figure 3.29. Map of active1 concentrated flow tracks (CFTs) at end of second growing season at Lewis site, Greene County in Georgia Piedmont. 1Active CFTs are scoured at least part way into the Streamside Management Zone.
113
Figure 3.30. Map of active1 concentrated flow tracks (CFTs) in a mature forest
five months post-burn, at Watson Springs site, Greene County, in Georgia Piedmont.
1Active CFTs are scoured at least part way into the Streamside Management Zone.
114
Figure 3.31. Map of active1 concentrated flow tracks (CFTs) in a mature forest
14 months post-burn, at Watson Springs site, Greene County, in Georgia Piedmont.
1Active CFTs are scoured at least part way into the Streamside Management Zone.
115
CHAPTER 4
CONCLUSIONS
The runoff cup method described appeared to be a functional method of studying
runoff frequency. Overland flow occurred on the two forest regeneration areas studied,
reaching the upper edge of the Streamside Management Zone (SMZ). The mean
frequency of overland flow for storms over 13 mm during the first nine months following
planting was 20.3% at the Lewis site and 8.7% at the Vanir site. This decreased to 10.9%
at Lewis and 4.7% at Vanir, over the next 11 months. The combined plantation mean
runoff frequencies were 13.0% the first period and 7.4% the second period, for an
average of 9.7% for the 20 month study period. The highest response rate for any single
storm was 77.1% at Lewis during the first growing season. Overland flow occurred in at
least one plantation cup in all storms over 13 mm in depth in the first period and 98.0%
of these storms in the second period. It was also widespread spatially around the edge of
the SMZ, occurring at least once in 83.3% of the cups in the first period, and 74.5% the
second period. Mean runoff frequency at Lewis differed statistically from that at Vanir,
for the combined dormant seasons and during the entire study period.
Overland flow also occurred on the mature forest reference site at 2.2%
frequency, in 79.2% storms, and at least once in 64.6% of the cups. The highest cup
response rate of any single storm was 38.0%. Mean runoff frequency at the reference site
differed statistically from the Lewis plantation, for the combined dormant seasons, the
combined growing seasons, and the entire study period. Mean runoff frequency at the
116
reference site differed statistically from the combined plantations for only the entire study
period. Mean runoff frequency at the reference site was statistically similar to the Vanir
plantation.
The best rainfall predictors of frequency of overland flow were: the “R” factor of
the Revised Universal Soil Loss Equation (RUSLE), storm depth, and maximum 24-hour
and 6-hour intensities, with linear regression r2 values between 0.46 and 0.75. Bare
ground decreased from a first year value of 31% at Lewis, 27% at Vanir, and 29% for the
combined pantations, to second year values of 27% at Lewis, 11% at Vanir, and 22% for
the combined plantations. Litter depth increased from two mm after one growing season
to four mm after the second growing season from the combined plantations. Bare ground
was 29% and litter depth 5mm in the reference stand in the second period of the study.
Dissolved nitrate means were 2.1 the first year and 0.8 mg/L the second year from
storm grab samples of combined plantations. Dissolved phosphate means were 0.23 the
first year and 0.12 mg/L the second year. Total Suspended Solid (TSS) means were 54
the first year and 36 mg/L the second year from grab samples from combined plantations.
The TSS mean from overland flow collectors was 66 mg/L the second growing season
from only one plantation. The highest grab-sample values were >6.0 for nitrates, 1.85 for
phosphates, and 1436 for TSS, all in mg/L.
Active concentrated flow tracks that penetrated the SMZ averaged 46 m (151 ft)
long in the management area, 0.9 m (3. ft) deep, and had a hydrologic contributing area
of 0.3 ha (0.7 ac). Active concentrated flow tracks that infiltrated in the SMZ averaged
44 m (144 ft) long in the management area, 0.6 m (2 ft ) deep and had a hydrologic
contributing area of 0.2 ha (0.5 ac). Most of these CFTs followed old agricultural gullies
117
and were therefore deeper than new CFTs . Concentrated flow traveled up to 146 m (479
ft) in the management area and up to 26 m (85 ft) in the SMZ, in this study.
Timber managers interested in water quality should pay more attention to
sensitive areas such as variable source areas, concentrated flow tracks, and old
agricultural gullies. These could be factored into decisions about placements of roads,
ditches, decks, site preparation, and streamside management zones and timing of
management activities.
Further research should be done on surface runoff generation in forests in general.
More specific questions during the regeneration window pertain to the roles of
topography, microtopography, and compaction.
118
CITATIONS
Abu-Zreig, M., R.P. Rodra, and H.R. Whitely. 2001. Validation of a vegetated filter strip model (VFSMOD). Hydrological Processes 15:729-742.
Agassi, M., I. Shainberg, and J. Morin. 1981. Effect of electrolyte concentration and soil sodicity on infiltration rate and crust formation. Soil Science Society of America Journal 45:848-851.
Allan, C.J., and N.T. Roulet. 1994. Runoff generation in zero-order precambrian shield catchments: The stormflow response of a heterogeneous landscape. Hydrological Processes 8: 369-388
Amerman, C.R. 1965. The use of unit-source watershed data for runoff prediction. Water Resources Research 1:499-507.
Atkinson, T.C. 1978. Techniques for measuring subsurface flow on hillslopes, p. 73-120, In M. J. Kirkby, ed. Hillslope Hydrology. John Wiley and Sons, Chichester, England.
Barling, R.D., and I.D. Moore. 1994. Role of buffer strips in management of waterway pollution - a review. Environmental Management 18:543-558.
Barnes, K.H. 1998. The effects of sedimentation on Georgia's fish assemblages with emphasis on the Upper Etowah system. MS, University of Georgia, Athens, Ga.
Beasley, R.S. 1979. Intensive site preparation and sediment losses on steep watersheds in the Gulf coastal Plain. Soil Science Society of America Journal 43:412-417.
Beasley, R.S., and A.B. Granillo. 1988. Sediment and water yields from managed forests on flat Coastal Plain sites. Water Research Bulletin 24:361-366.
Beasley, R.S., A.B. Granillo, and V. Zillmer. 1986. Sediment losses from forest management: Mechanical vs. chemical site preparation. Journal of Environmental Quality 15:413-416.
Betson, R.P. 1964. What is watershed runoff? Journal of Geophysical Research 69:1541-1552.
Betson, R.P., and J.B. Marius. 1969. Source areas of storm runoff. Water Resources Research 5:574-582.
119
Beven, K., M.J. Kirkby, N. Schofield, and A.F. Tagg. 1984. Testing a physically-based flood forescasting model (TOPMODEL) for three U.K. catchments. Journal of Hydrology 69:119-143.
Blackburn, W., J.C. Wood, and M.G. DeHaven. 1986. Storm flow and sediment losses from site-prepared forest land in East Texas. Water Resources Research 22:776-784.
Blackburn, W.H., and J.C. Wood. 1990. Nutrient export in stormflow following forest harvesting and site-preparation in East Texas. Journal of Environmental Quality 19:402-408.
Blinn, C.R., and M.A. Kilgore. 2001. Riparian management practices - A summary of state guidelines. Journal of Forestry 99:11-17.
Bonnell, M., and D.A. Gilmour. 1978. The development of overland flow in a tropical rainforest catchment. Journal of Hydrology 39:365-382.
Brady, N.C., and R.R. Weil. 1999. The Nature and Property of Soils. 12 ed. Prentice-Hall Inc., Upper Saddle River, NJ. pp. 831
Burt, T.P. 1989. Storm runoff generation in small catchments in relation to the flood response of large basins, p. 11-35, In K. Beven and P. Carling, eds. Floods-Hydrological, sedimentological, and geomorphological implications. J. Wiley, New York.
Castillo, V.M., A. Gomez-Plaza, and M. Martinez-Mena. 2003. The role of antecedent soil water content in the runoff response of semi-arid catchments: a simulation approach. Journal of Hydrology 284:114-130.
Cavalcanti, C.G., and B.G. Lockaby. 2005. Effects of sediment deposition on fine root dynamics in riparian forests. Soil Science Society of America Journal 69:729-737.
de Pinho, A.P. 2003. Retancao de atrazina e picloram, provenientes de escoamento superficial, em zonas riparias. Ph.D. Dissertation, Universidada Federal de Vicosa, Minais Gerais, Brasil.
DeBano, L.F. 2000. The role of fire and soil heating on water repellancy in wildland environments: a review. Journal of Hydrology 231/232:195-206.
Dietrich, W.E., D.M. Windsor, and T. Dunne. 1982. Geology, climate, and hydrology of Barro Colorado Island, p. 27-46, In E. G. J. Leigh, et al., eds. Seasonal Rhythms and the Ecology of a Tropical Forest: Seasonal Rhythms and Long-term Changes. Smithsonian Institution Press, Washington, D.C.
120
Douglass, J.E., and D.H. Van Lear. 1983. Prescribed burning and water quality at ephemeral streams in the piedmont of South Carolina. Forest Science 29:181-189.
Dunne, T., and R.D. Black. 1970a. An experimental investigation of runoff production in permeable soils. Water Resources Research 6:478-490.
Dunne, T., and R.D. Black. 1970b. Partial area contributions to storm runoff in a small New England watershed. Water Resources Research 6:1296-1311.
Dunne, T., and L.B. Leopold. 1978. Water in Environmental Planning Will Freeman and Co., New York.
Dunne, T., T.R. Moore, and C.H. Taylor. 1975. Recognition and prediction of runoff-producing zones in humid regions. Hydrological Sciences Bulletin 20:305-325.
Dykes, A.P. 1997. Rainfall interception from a lowland tropical forest in Brunei. Journal of Hydrology 200:260-279.
Dykes, A.P., and J.B. Thornes. 2000. Hillslope hydrology in tropical rainforest steeplands in Brunei. Hydrological Processes 14:215-235.
Eaton, A.D., L.N. Clesceri, A.E. Greenberg, and M.A.H. Franson, (eds.) 1995. Standard methods for the examination of water and wastewater, 19th ed. American Public Health Association, Washington, D.C.
Elsenbeer, H., and R.A. Vertessy. 2000. Stormflow generation and flowpath characteristics in an Amazonian rainforest catchment. Hydrological Processes 14:2367-2381.
Ferreira, A.J.D., C.O.A. Coelho, R.P.D. Walsh, R.A. Shakesby, A. Ceballos, and S.H. Doerr. 2000. Hydrological implications of soil water-repellency in Eucalyptus globulus forests, north-central Portugal. Journal of Hydrology 231-232:165-177.
Field, J.P., K.W. Farrish, and E.A. Carter. 2003. Soil and nutrient losses following site preparation burning in a harvested loblolly pine site. Transactions of the ASAE 46:1697-1703.
Field, J.P., K.W. Farrish, B.P. Oswald, M.T. Ramig, and E.A. Carter. 2005. Forest site preparation effects on soil and nutrient losses in East Texas. Transactions of the ASAE 48:861-869.
Findeling, A., S. Ruy, and E. Scopel. 2003. Modeling the effects of a partial residue mulch on runoff using a physically based approach. Journal of Hydrology 275:49-66.
121
Fox, D.M., and R.B. Bryan. 2000. The relationship of soil loss by interrill erosion to slope gradient. Catena 38:211-222.
Fox, T.R., J.A. Burger, and R.E. Kreh. 1986. Effects of site preparation on nitrogen dynamics in the Southern Piedmont. Forest Ecology and Management 15:241-256.
Franklin, D.H., M.L. Cabrera, J.L. Steiner, D.M. Endale, and W.P. Miller. 2001. Evaluation of percent flow captured by a small infield runoff collector. Transactions of the ASAE. 44:551-554.
Gabet, E.J., and T. Dunne. 2003. Sediment detachment by rain power. Water Resources Research 39:1002, 1-12.
Georgia Environmental Protection Division, and Georgia Forestry Commission. 1999. Georgia's Best Management Practices for Forestry, Macon, GA. pp.66.
Georgia GIS Clearinghouse. [Online] http://gis1.state.ga.us Accessed Oct 2004.
Georgia State Climatology Office. 1997. Monthly precipitation: 30 year Averages [Online] http://climate.engr.uga.edu/plant_sciences/precip.html Accessed July 2004 Georgia State Climatology Office. 2002. Daily summaries. 1990-1999 [Online] http://climate.engr.uga.edu/plant_sciences/daily_90s.html Accessed July 2004 Georgia State Climatology Office. 2003. Daily summaries. 2000-2003 [Online] http://climate.engr.uga.edu/plant_sciences/daily_2000s.html Accessed July 2004 Germann, P.F. 1986. Rapid drainage response to precipitation. Hydrological Processes
1:3-14.
Ghadiri, H., C.W. Rose, and W.L. Hogarth. 2001. The influence of grass and porous buffer strips on runoff hydrology and sediment transport. Transactions of the ASAE 44:259-268.
Godsey, S., H. Elsenbeer, and R. Stallard. 2004. Overland flow generation in two lithologically distinct rainforest catchments. Journal of Hydrology 295:276-290.
Grace, J.M.I. 2004. Soil erosion following forest operations in the southern Piedmont of central Alabama. Journal of Soil and Water Conservation 59:160-166.
122
Han, J., J.S. Wu, and C. Allan. 2005. Suspended sediment removal by vegetative filter strips treating highway runoff. Journal of Environmental Science and Health Part A- Toxic/Hazardous Substances and Environmental Engineering 40:1637-1649.
Hansen, W.F. 2001. Identifying stream types and management implications. Forest Ecology and Management 143:39-46.
Heede, B.H. 1987. Overland flow and sediment delivery five years after timber harvest in a mixed conifer forest, Arizona, U.S.A. Journal of Hydrology 91.
Herwitz, S.R. 1986. Infiltration-excess caused by stemflow in a cyclone-prone tropical rainforest. Earth Surface Processes and Landforms 11:401-412.
Hewlett, J.D. 1978. Forest water quality: an experiment in harvesting and regenerating Piedmont Forest Land. School of Forest Resources, University of Georgia, Athens, GA.
Hewlett, J.D., and A.R. Hibbert. 1963. Moisture and energy conditions within a sloping
soil mass during drainage. Journal of Geophysical Research 68:1081-1087. Hewlett, J.D., H.E. Post, and R. Doss. 1984. Effect of clearcut silviculture on dissolved
ion export and water yield in the Piedmont. Water Resources Research 20:1030-1038.
Hewlett, J.D., and A.R. Hibbert. 1967. Factors affecting the response of small watersheds to precipitation in humid areas, p. 275-290, In W. E. Sopper and H. W. Lull, eds. Forest Hydrology. Pergamon Press, New York.
Horton, R.E. 1933. The role of infiltration in the hydrologic cycle. Transactions of the American Geophysical Union 14:446-460.
Horton, R.E. 1940. An approach towards a physical interpretation of infiltration-capacity. Soil Science Society Proceedings 5?:399-417.
Hunzinger, H. 1997. Hydrology of montane forests in the Sierra de San Javier, Tucuman, Argentina. Mountain Research and Development 17:299-308.
Ice, G.R., G.W. Stuart, J.B. Waide, L.C. Irland, and P.V. Ellefson. 1997. Twenty five years of the Clean Water Act: How clean are forest practices? Journal of Forestry 95:9-13.
Jackson, C.R. In press. Wetland hydrology, In D. P. Batzer and R. Sharitz, eds. Ecology of Freshwater and Marine Wetlands. University of California Press.
123
Jayawardena, A.W., and R.R. Bhuiyan. 1999. Evaluation of an interrill soil erosion model using laboratory catchment data. Hydrological Processes 13:89-100.
Johansen, M.P., T.E. Hakonson, and D.D. Breshears. 2001. Post-fire runoff and erosion from rainfall simulation: contrasting forests with shrublands and grasslands. Hydrological Processes 15:2953-2965.
Keim, R.F., and S. Schoenholtz. 1999. Functions and effectiveness of silvicultural streamside management zones in loessial bluff forests. Forest Ecology and Management 118:197-209.
Kirkby, M.J., and R.J. Chorley. 1967. Throughflow, overland flow, and erosion. International Association of Scientific Hydrology Bulletin 12:5-21.
Kothyari, B.P., P.K. Verma, B.K. Joshi, and U.C. Kothyari. 2004. Rainfall-runoff-soil and nutrient loss relationships for plot size areas of bhetagad watershed in Central Himalaya, India. Journal of Hydrology 293:137-150.
Lacey, S.T. 2000. Runoff and sediment attenuation by undisturbed and lightly disturbed forest buffers. Water, Air and Soil Pollution 122:121-138.
Lesack, L.F.W. 1993. Water-balance and hydrologic characteristics of a rain-forest catchment in the cetral Amazon basin. Water Resources Reseach 29:759-773.
Liebenow, A.M., W.J. Elliot, J.M. Laflen, and K.D. Kohl. 1990. Interrill erodibility: Collection and analysis of data from cropland soils. Transactions of the ASAE 33:1182-1888.
Malmer, A. 1996. Hydrological effects and nutrient losses of forest plantation establishment on tropical rainforest land in Sabah, Malaysia. Journal of Hydrology 174:129-148.
Martinez-Mena, M., V.M. Castillo, and J. Albaladejo. 2001. Hydrological and erosional response to natural rainfall in a semi-arid area of south-east Spain. Hydrological Processes 15:557-571.
Miller, W.P., R. Jackson, and T.C. Rasmussen. 1999. Readings in soils and hydrology. A text for CRSS/FORS 3060 University of Georgia, Athens, GA. pp. 170
McClurkin, D.C., P.D. Duffy, and N.S. Nelson. 1987. Changes in forest floor and water
quality following thinning and clearcutting of 20-year old pine. Journal of Environmental Quality 16:237-241.
124
McClurkin, D.C., P.D. Duffy, S.J. Ursic, and N.S. Nelson. 1985. Water quality effects of clearcutting upper Coastal Plain loblolly pine plantations. Journal of Environmental Quality 14:329-332.
McDowell, T.R., and J.M. Omernik. 1977. Nonpoint source-stream nutrient level relationships: A nationwide study.Supplement 1: Nutrient map reliability EPA-600/3-77-105. US Environmental Protection Agency, Corvallis, OR.
McGregor, K.C., R.L. Bingner, A.J. Bowie, and G.R. Foster. 1995. Erosivity index values for northern Mississippi. Transactions of the ASAE 38:1037-1047.
Meding, S.M., L.A. Morris, C.A. Hoover, W.L. Nutter, and M.L. Cabrera. 2001. Denitrification of a long-term forested land treatment system in the piedmont of Georgia. Journal of Environmental Quality 30:1411-1420.
Miller, E.L. 1984. Sediment yield and storm flow response to clear-cut harvest and site preparation in the Ouachita Mountains. Water Resources Research 20:471-475.
NCASI. 1994. Forests as Nonpoint Sources of Pollution, and Effectiveness of Best Management Practices Technical Bulletin 672. National Council of the Paper Industry for Air and Stream Improvement, Inc, New York.
Nutter, W.L. 1973. The role of soil water in the hydrologic behavior of upland basins, p. 181-193, In M. Stelly, ed. Field Soil Water Regime. Soil Science Society of America, Madison, WI.
Onset Computer Corporation. 2001. Data Logging Rain Gauge Manual RG2 and RG2-M Onset Computer Corporation, Bourne, MA.
Patric, J.H. 1980. Effects of wood products harvest on forest soil and water relations. Journal of Environmental Quality 9:73-79.
Patric, J.H., J.O. Evans, and J.D. Helvey. 1984. Summary of sediment yield data from forested land in the United States. Journal of Forestry 82:101-104.
Pearce, P.A., G.W. Frasier, M.J. Trlica, W.C. Leininger, J.D. Stednick, and J.L. Smith. 1998. Sediment filtration in a montane riparian zone under simulated rainfall. Journal of Range Management 51:309-314.
Peterjohn, W.T., and D.L. Correll. 1984. Nutrient dynamics in an agricultural watershed: Observations on the role of a riparian forest. Ecology 65:1466-1475.
Phillips, J.D. 1989. An evaluation of the factors determining the effectiveness of water quality II. Runoff processes. Journal of Hydrology 107:133-145.
125
Pilgrim, D.H., D.D. Huff, and T.D. Steele. 1978. A field evaluation of subsurface and surface runoff. Journal of Hydrology 38:319-341.
Ponce, V.M., and R.H. Hawkins. 1996. Runoff curve number: Has it reached maturity? Journal of Hydrologic Engineering 1:11-19.
Pye, J.M., and P.M. Vitousek. 1985. Soil and nutrient removals by erosion and windrowing at a southeastern U.S. piedmont site. Forest Ecology and Management 11:145-155.
Quansah, C. 1985. The effect of soil type, slope, flow rate and their interactions on detachment by overland flow with and without rain, p. 19-28, In P. D. Jungerius, ed. Soils and Geomorphology, Catena Supp. 6.
Ragan, R.M. 1967. An experimental investigation of partial area contributions, p. 241-251 Hydrological Aspects of the Utilization of Water, Vol. Publication 76. International Association of Scientific Hydrology (IAHS), Gentbrugge, Belgium.
Richter, D.D., and D. Markewitz. 2001. Understanding soil change. Soil sustainability over millenia, centuries, and decades Cambridge University Press, New York.pp. 255.
Rivenbark, B.L., and C.R. Jackson. 2004. Concentrated flow breakthroughs moving through silvicultural streamside management zones: Southeastern Piedmont, USA. Journal of American Water Resources Association 40:1043-1052.
Robichaud, P.R., and T.A. Waldrop. 1994. A comparison of surface runoff and sediment yields from low-severity and high-severity site preparation burns. Water Resources Bulletin 30:27-34.
Roper, D.M. 1996. Boyhood memories of Watson Springs resort. North Georgia History Spring:8-11.
Schmitt, T.J., M.G. Dosskey, and K.D. Hoagland. 1999. Filter strip performance and processes for different vegetation, widths, and contaminants. Journal of Environmental Quality 28:1479-1489.
Schreiber, J.D., P.D. Duffy, and D.C. McClurkin. 1980. Aqueous sediment-phase nitrogen yields from five southern pine watersheds. Soil Science Society of America Journal 44:401-407.
Sheridan, J.M., R.R. Lowrance, and H.H. Henry. 1996. Surface flow sampler for riparian studies. Applied Engineering in Agriculture 12:183-188.
Shainberg, I., A.I. Mamedov, and G.J. Levy. 2003. Role of wetting rate and rain energy in seal formation and erosion. Soil Science 168:54-62.
126
Sharma, P.P., S.C. Gupta, and G.R. Foster. 1995. Raindrop-induced soil detachment and sediment transport from interrill areas. Soil Science Society of America Journal 59:727-734.
Sidle, R.C., Y. Tsuboyama, S. Noguchi, I. Hosoda, M. Fujieda, and T. Shimizu. 2000. Stormflow generation in steep forested headwaters: A linked hydrogeomorphic paradigm. Hydrological Processes 14:369-385.
Sinun, W., W.W. Meng, I. Douglas, and T. Spencer. 1992. Throughfall, stemflow, overland-flow and throughflow in the Ulu Segama rain forest, Sabah Malaysia. Philosophical Transactions of the Royal Society of London Series B - Biological Sciences 335:389-395.
Srivastava, P., D.R. Edwards, T.C. Daniel, P.A.J. Moore, and T.A. Costello. 1996. Performance of vegatative filter strips with varying pollutant source and filter strip lengths. Transactions of the ASAE 39:2231-2239.
Stomph, T.J., N. de Ridder, T.S. Steenhuis, and N.C. van de Giesen. 2002. Scale effects of Hortonian overland flow and rainfall-runoff dynamics laboratory validation of a process-based model. Earth Surface Processes and Landforms 27:847-855.
Sutherland, A.B., J.L. Meyer, and E.P. Gardiner. 2002. Effects of land cover on sediment regime and fish assemblage structure in four southern Appalachian streams. Freshwater Biology 47:1791-1805.
Swift, L.W., Jr. 1986. Filter strip widths for forest roads in the southern Appalachians. Southern Journal of Applied Forestry 10:27-34.
Teixera, P.C., and R.K. Misra. 1997. Erosion and sediment characteristics of cultivated forest soils as affected by the mechanical stability of aggregates. Catena 30:119-134.
Tennessee Valley Authority. 1965. Area-stream factor correlation, a pilot study in the Elk River basin. Bulletin of International Scientific Hydrology 10:22-37.
Thompson, M.T. 1998. Forest statistics for Georgia, 1997 [Online]. Available by USDA Forest Service, Southern Research Station http://www.srs.fs.usda.gov/pubs (verified 10/2002).
Trimble, S.W. 1974. Man-induced soil erosion on the Southern Piedmont 1700-1970 Soil Conservation Society of America, Ankenny, IA, pp.180.
Ursic, S.J. 1991. Hydrologic effects of two methods of harvesting mature southern pine. Water Resources Bulletin 27:303-315.
127
US Environmental Protection Agency. 2000. Ambient water quality criteria recommendations. Information supporting the development of state and tribal nutrient criteria for rivers and streams in nutrient ecoregion IX [Online] http://www.epa.gov/waterscience/criteria/nutrient/ecoregions/rivers/rivers_9.pdf (posted Dec 2000; verified 03/30/2006).
US Environmental Protection Agency. 2002. National water quality inventory report to Congress (305(b) report). 2002 National Assessment Database [Online] http://www.epa.gov/305b; verified 11/2005.
United States Geological Survey. Surface-water data for the nation. Gauge # 02218300 [Online] http://waterdata.usgs.gov/nwis/sw. Accessed 2005.
van de Giesen, N.C., T.J. Stomph, and N. de Ridder. 2000. Scale effects of Hortonian overland flow and rainfall- runoff dynamics in a West African catena landscape. Hydrological Processes 14:165-175.
Van Lear, D.H., J.E. Douglass, S.K. Cox, and M.K. Augspurger. 1985. Sediment and nutrient export from burned and harvested pine watersheds in the South Carolina Piedmont. Journal of Environmental Quality 14:169-174.
Waldron, L.J., and S. Dakessian. 1982. Effect of grass, legume, and tree roots on soil shearing resistance. Soil Science Society of America Journal 46:894-899.
Wallach, R., and D. Zaslavsky. 1991. Lateral flow in a layered profile of an infinite uniform slope. Water Resources Reseach 27:1809-1818.
Ward, J.M., and C.R. Jackson. 2004. Sediment trapping within forestry streamside management zones: Georgia Piedmont, USA. Journal of American Water Resources Association 40:1421-1431.
Weyman, D.R. 1970. Throughflow on hillslopes and its relation to the stream hydrograph. Bulletin of International Association of Scientific Hydrology 15:25-33.
White, W.J. 2003. Retention of sediment and phosphorus in forested streamside management zones of the Georgia Piedmont under simulated overland flow conditions. MS, University of Georgia, Athens, GA.
Wigington, P.J., Jr., T.J. Moser, and D.R. Lindeman. 2005. Stream network expansion: A riparian water quality factor. Hydrological Processes 19:1715-1721.
Wischmeier, W.H., and D.D. Smith. 1958. Rainfall energy and its relationship to soil loss. Transactions of the American Geophysical Union 39:285-291.
128
Wohlgemuth, P.M., K.R. Hubbert, and P.R. Robichaud. 2001. The effects of log erosion barriers on post-fire hydrologic response and sediment yield in small forested watersheds, southern California. Hydrological Processes 15:3053-3066.
Zhang, X.C., and M.A. Shao. 2003. Effects of vegetation coverage and management practices on soil nitrogen loss by erosion in a hilly region of the Loess plateau in China. Acta Botanica Sinica 45:1195-1203.
Ziegler, A.D., T.W. Giambelluca, L.T. Tran, T.T. Vana, M.A. Nullet, J. Fox, T.D. Vien, J. Pinthong, J.F. Maxwell, and S. Evett. 2004. Hydrological consequences of landscape fragmentation in mountainous northern Vietnam: evidence of accelerated overland flow generation. Journal of Hydrology 287:124-146.